Skip to main content

Full text of "Atoll research bulletin"

See other formats




^^^m 



, ■ 



ill 






Ibb 



■ ■ ■ i 



Nos. 282-292 



May 1985 






ATOLL RESEARCH 
BULLETIN 



282. Feral cats on Jarvis Island: Their effects and 
their eradication 

by Mark J. Rauzon 

283. Vegetation and flora of Nui Atoll, Tuvalu 
by C. D. Woodroffe 

284. Initial recolonization of Funafuti Atoll coral 
reefs devastated by hurricane "Bebe" 

by Hans Mergner 

285. Status and ecology of marine turtles at 
Johnston Atoll 

by George H. Balazs 

286. Environmental survey ofMataiva Atoll, 
Tuamotu Archipelago, French Polynesia 
by B. Delesalle and Colleagues 

287. Checklist of the vascular plants of the 
Northern Line Islands 

by Lyndon Wester 



288. Community structure of reef-building corals in 
the Florida Keys: Carysfort Reef, Key Largo 
and Long Key Reef, Dry Tortugas 

by Phillip Dustan 

289. The distribution, abundance and primary 
productivity of submerged macrophytes in a 
Belize barrier-reef mangrove system 

by Mark M. Littler, Phillip R. Taylor, Diane 
S. Littler, Robert H. Sims and James N. 
Norris 

290. Some observations on Nesillas aldabranus, the 
endangered brush warbler of Aldabra Atoll, 
with hypotheses on its distribution 

by C. Hambler, K. Hambler and J. M. 
Newing 

291. Changes in the distribution of the coccid 
Icerya seychellarum Westw. on Aldabra Atoll 
in relation to vegetation density 

by D. McC. Newbery and M. G. Hill 

292. Short original articles 
by Various Authors 




Issued by 

THE SMITHSONIAN INSTITUTION 

Washington, D.C., U.S.A. 



ATOLL RESEARCH BULLETIN 

MAY 1985 

Nos. 282-292 



282. Feral cats on Jarvis Island: Their effects and their 
eradication 

by Mark J. Rauzon 

283. Vegetation and flora of Nui Atoll, Tuvalu 
by C. D. Woodroffe 

284. Initial recolonization of Funafuti Atoll coral reefs 
devastated by hurricane "Bebe" 

by Hans Mergner 

285. Status and ecology of marine turtles at Johnston Atoll 
by George H. Balazs 

286. Environmental survey of Mataiva Atoll, Tuamotu Archipelago, 
French Polynesia 

by B. Delesalle and Colleagues 

287. Checklist of the vascular plants of the Northern Line 
Islands 

by Lyndon Wester 

288. Community structure of reef-building corals in the Florida 
Keys: Carysfort Reef, Key Largo and Long Key Reef, Dry 
Tortugas 

by Phillip Dustan 

289. The distribution, abundance and primary productivity of 
submerged macrophytes in a Belize barrier-reef mangrove 
system 

by Mark M. Littler, Phillip R. Taylor, Diane S. Littler, 
Robert H. Sims and James N. Norris 

290. Some observations on Nesillas aldabranus , the endangered 
brush warbler of Aldabra Atoll, with hypotheses on its 
distribution 

by C. Hambler, K. Hambler and J. M. Newing 

291. Changes in the distribution of the coccid Icerya 
seychel larum Westw. on Aldabra Atoll in relation to 
vegetation density 

by D. McC. Newbery and M. G. Hill 

292. Short original articles 
by Various Authors 



Issued by 
THE SMITHSONIAN INSTITUTION 
Washington, D.C. , U.S.A. 



ACKNOWLEDGMENT 



The Atoll Research Bulletin is issued by the Smithsonian 
Institution, as a part of its activity in tropical biology, to place 
on record information on the biota of tropical islands and reefs, 
and on the environment that supports the biota. The Bulletin is 
supported by the National Museum of Natural History and is produced 
and distributed by the Smithsonian Press. This issue has been par- 
tially financed with funds contributed by readers and authors. The 
editing is done by members of the Museum staff and by Dr. D. R. 
Stoddart. 

The Bulletin was founded and the first 117 numbers issued by the 
Pacific Science Board, National Academy of Sciences, with financial 
support from the Office of Naval Research. Its pages were largely 
devoted to reports resulting from the Pacific Science Board's Coral 
Atoll Program. 



The sole responsibility for all statements made by authors of 
papers in the Atoll Research Bulletin rests with them, and statements 
made in the Bulletin do not necessarily represent the views of the 
Smithsonian nor those of the editors of the Bulletin. 



Articles submitted for publication in the Atoll Research Bulletin 
should be original papers in a format similar to that found in recent 
issues of the Bulletin. Manuscripts should be typewritten and double 
spaced. After the manuscript has been reviewed and accepted, the 
author will be provided with a page format with which to prepare a 
camera-ready copy of the manuscript. 



Editors 

F. R. Fosberg 
Ian G. Macintyre 
M.-H. Sachet 

Smithsonian Institution 
Washington, D.C. 20560 

D. R. Stoddart 

Department of Geography 
University of Cambridge 
Downing Place 
Cambridge, England 




Dr. Coolidge receiving the Edward W. Browning Achievement Award 
from S. Dillon Ripley in 1978 



Harold Jefferson Coolidge 1904-1985 

It saddens us to have to report the death, on February 15, 1985, 
of Dr. Harold Jefferson Coolidge, whose support as Director of the 
Pacific Science Board, National Academy of Sciences-National Research 
Council, was ultimately responsible for the Board's Coral Atoll Program 
and for the founding and first 15 years of publication of the Atoll Re- 
search Bulletin. 

Hal was born on January 15, 1904, in Boston, Massachusetts. He 
received his higher education from Harvard University, and continued 
his association with Harvard as assistant curator at its Museum of 
Comparative Zoology from 1929-46. As a professional mammalogist he 
took part in several expeditions, among them one crossing Central Africa 
with the Harvard Medical Team in 1926-27, leading the Indo-China Divi- 
sion, Kelley-Roosevelt ' s Field Museum Expedition 1928-29, and organizing 
and leading the Harvard Asiatic Primate Expedition, 1937. His work at 
the Museum resulted in many scientific articles and monographs on the 
Genus Gorilla, Indo-Chinese Forest Ox or Kouprey, Pygmy Chimpanzee and 
other primates and mammals. 

However, his major reputation was as an enormously successful pro- 
moter of research in all fields in the Pacific Basin and especially the 
islands, and as a leading international conservationist. Together with 
other scientists interested in the Pacific islands he helped to organize 
the National Research Council's conference in 1946 that resulted in the 
establishment of the Pacific Science Board. As the Board's Director 
from 1946 to 1970 he was responsible for such major research programs 
as the Coordinated Investigations in Micronesian Anthropology (CIMA) , 
the Scientific Investigations in Micronesia (SIM) , and the Coral Atoll 
Program. He was a major figure in the Pacific Science Association, and 
served as Secretary General of the Tenth Pacific Science Congress, held 
in Honolulu in 1961. 

In the field of conservation, Hal played a major role in the found- 
ing of the International Union for the Conservation of Nature and Natural 
Resources in 1948, and was principally responsible for bringing the 
Second Technical Meeting of this new organization to the U.S., at Lake 
Success, in 1949. He served the Union in many official capacities over 
the years as Vice President, Chairman of the Survival Service Commission 
and the International Commission on National Parks, and as President from 
1966 to 1972. He was named Honorary President of the Union in 1972. 

Hal was the recipient of three Honorary Doctor of Science degrees, 
and many awards for his international conservation activities, some of 
which were the J. Paul Getty Wildlife Conservation Prize, the Edward W. 
Browning Achievement Award, John C. Phillips Medal, and the Gold Medal 
of the New York Zoological Society. 

Our close association with him extended through almost the entire 
24 years of the Pacific Science Board. In 1949, in response to a re- 
quest by the South Pacific Commission for information on coral atoll 



11 

ecology, he held a small informal meeting at which the idea of a coral 
atoll research project was put forward. Coolidge formalized this idea 
into a proposal to the Office of Naval Research, and it was funded and 
carried on for over five years during which time major expeditions were 
sent to five representative atolls in the Pacific. The Atoll Research 
Bulletin was started in 1951 to make widely available the results of 
these expeditions and related investigations and studies of coral atolls 
and reefs. The Coral Atoll Program gave a principal impetus to the 
enormous amount of work that has been done on coral atolls and reefs 
since and is still going on under various auspices. This honor is, of 
course, shared with the U.S. Geological Survey, working on Bikini and 
other northern Marshall Atolls before and after the atomic bomb tests. 
Much of our own work was carried out under informal joint sponsorship 
of the Pacific Science Board and the Geological Survey. The prestige 
of the Board and its Director drew in grant support that made the non- 
geological aspects of the atoll and reef research possible. Coolidge' s 
enthusiastic backing for this was unfailing and stimulated the activi- 
ties of the many participants in the program. Its effects have continued 
and will be productive under many guises as time and coral reef research 
go on. 

Now, at its end, we salute our friend , and his career of many and 
varied accomplishments. His name will long be remembered, especially 
in the context of research in the Pacific Ocean area, and of conservation 
on a global scale. 



ATOLL REStARCH BULLETIN 
No. 282 



FERAL CATS ON JARVIS ISLAND: 
THEIR EFFECTS AND THEIR ERADICATION 



BY 

Iark J- Rauzon 



Issued By 

THE SMITHSONIAN INSTITUTION 

Washington, D-C., U.S.A. 

May 1985 



u 



r 



C£ 



< 

z 

O 
< 



< 

> 

< 




CD 

o 



en 



i 

<-> 

O 

z 
o 

M 

S3 

u 
o 



FERAL CATS ON JARVIS ISLAND: 
THEIR EFFECTS AND THEIR ERADICATION 



RY 



Mark J. Rauzon* 



INTRODUCTION 



Island ecosystems have proven to be particularly sensitive to 
human disturbances (Bourne 1975; Byrne 1980; Jarvis 1979). This was 
first noted by Charles Darwin in the explanation of his theory of 
natural selection (Byrne 1980). Since then, qualities of insular 
species have been examined by various authors in an attempt to 
understand the basis for island vulnerability. 

Within island environments, there is generally less competition 
and predation than there is in corresponding continental habitats 
(Jarvis 1979). Introduced species are often competitively superior to 
insular species. However, the success of introduced species on oceanic 
islands can be attributed, in part, to human disturbance. The initial 
advantage of introduced species is their ability to withstand the types 
of disturbances associated with man (Egler 1942; Mueller-Dombois and 
Spatz 1972). 

Of all the environmental changes caused by introduced species, 
predation has one of the most immediate effects on indigenous 
populations. Specifically, the feral cat ( Felis catus ) has played a 
major role in the eradication of native birds on islands. In New 
Zealand alone, these predators are implicated in the extinction of at 
least 6 endemic species and over 70 localized subspecies (Merton 1978). 



^University of Hawaii at Manoa, Department of Geography, Honolulu, 
Hawaii. 
Present address: P. 0. Box 4423, Berkeley, CA. 94704 



Research on Marion Island, a sub-Antarctic possession of South Africa, 
illustrates the potential magnitude of cat predation. By calculating 
the caloric needs of cats in all developmental stages and the caloric 
content of prey species, van Aarde (1980) was able to determine that 
200 petrel-sized birds were eaten per cat per year. At least 455,119 
birds had to be consumed to provide energy for the 2137 cats. Based on 
scat analysis of 375 cats on Macquarie Island, the yearly total of prey 
eaten are 56,000 rabbits ( Oryctolagus cuniculus ). 47,000 Antarctic 
prion ( Pachyptila desolata ) and 11,000 white-headed petrels 
( Pterodroma lessonii ) (Jones 1977). Cats in the Kerguelen 
Archipelago probably kill about 1.2 million birds every year (Pascal 
1980). 

Returning a disrupted ecosystem to a condition more closely 
resembling the original state requires the complete removal of 
introduced species, especially predators. Even so, unanticipated 
results can occur. Attempts to eradicate rats which were endangering 
the Bermuda petrel ( Pterodroma cahow ) resulted in an increase of 
tropicbirds ( Phaethon sp.) which colonized burrow nest sites formerly 
used by petrels. This problem was solved by fitting the burrows with 
baffles to exclude tropicbirds (Murphy 1964). 

Eradicating introduced animals, even from relatively small 
islands, is a difficult and exacting task. This is particularly true 
in dealing with carnivores, which are capable of ". . . learned 
behavioral responses" (Beck 1975). Feral cats were eliminated from 8 
islands in New Zealand. The complete removal of cats from Little 
Barrier Island spanned 4 years and involved 128 people and 3880 
man-days (Veitch pers. coram.). A partial reduction of cats from Marion 
Island, South Africa, was produced by the introduction of feline 
panleucopenia virus followed by mechanical control efforts (van Aarde 
pers. comm.). I report here on the tentative eradication of feral cats 
from Jarvis Island, central Pacific Ocean. 



JARVIS ISLAND 



Jarvis Island (0 22'S, 160 01'W) is a remote, emergent atoll 
located approximately 1300 miles south of Hawaii between the Line and 
Phoenix Island groups (Figure 1). The island is about 25 miles south 
of the equator and 200 miles southwest of the nearest island, Christmas 
Island, Kiribati. It consists of 1024 acres (about 1.6 square miles) 
of coral rubble, phosphatic guano and organic detritus (Bryan 1974; 
Hutchinson 1950). The desert-like climate is characteristic of atolls 
in the Equatorial Dry Zone which is bordered roughly by 5 latitude 
north and south. Slightly south of the equator and east of 180 
meridian lies the minimum rainfall region which includes Jarvis (Taylor 
1973). This scantily vegetated island is highly reflective of strong 
solar radiation which retards precipitation over the island even when 
rain is falling over the surrounding ocean (Christophersen 1927). Only 
30 inches of rain fell during the five-year period from 1975 to 1980. 
A remote automatic weather station was operated for about 3-1/2 years 
during that time (Vitousek, Kilonsky and Leslie 1980). To the north of 
the minimum rainfall zone, Christmas Island receives about 30 inches of 
rain per year. 

The scarce vegetation is typical of strand communities of the 
tropical Pacific and consists of Boerhaavia diffusa , Portulaca 
lutea , Sesuvium portulaca strum , Sida fallax , Tribulus cistoides , 
Lepturus repens , Eragrostis whitneyi and Abu ti Ion indicum 
(Christophersen 1927). The location and areal cover of major plant 
communities is shown in Figure 2. A similar map was reproduced by E. 
H. Bryan in 1942 and comparison between the 2 maps (Figure 3) indicate 
few changes except in the degree of Tribulus cover, a phenomenon 
which appears to be seasonal (Bryan 1942). 

The highest point of the island is the northwest beach crest which 
is approximately 25 feet above mean sea level. The topography 
indicates that Jarvis Island was once a horseshoe shaped atoll with a 
lagoon in the center. As the lagoon drained and filled, extensive beds 
of gypsum were formed (Hague 1862 in Hutchinson 1950; Jewell 1961). 
Continual deposition of seabird excreta created a valuable deposit of 
phosphatic guano. 

Jarvis Island was discovered by Captain Brown of the British ship 
ELIZA FRANCIS on 21 August 1821. Various ships visited this island, 
also known at the time as Jervis or Bunker, before 1856 when it was 
claimed by the American Guano Company and the United States Guano 



company under the U.S. Guano Act of 1856 which allowed ship captains to 
claim sovereignty over unoccupied islands. The U.S. government claimed 
that mere discovery did not give final title if not followed 
immediately by reasonable occupation. In February 1858, C. H. Judd 
took 23 Hawaiian laborers to begin mining guano (Judd 1960). 
Excavation of phosphatic guano lasted until 1879 (Bryan 1974). Initial 
estimates based on the vast numbers of seabirds present in 1856 
predicted that 7 million tons of guano were available. However, log 
reports from commercial guano vessels indicate it was unlikely that the 
output exceeded 12,000 tons annually. By the termination of the lease 
in 1879, approximately 300,000 tons were removed from Jarvis Island 
making it one of the richest deposits in the Central Pacific 
(Hutchinson 1950). The ownership of Jarvis was contested when Great 
Britain annexed the island in 1889 and leased the deposits to the 
Pacific Phosphate Company of London and Melbourne. Because so little 
high quality guano remained, the lease was allowed to expire (Bryan 
1974). Today, rows of low-grade spoil remain in 3 to 7 foot walls in 
the interior of the island providing dens for cats. 

The value of Jarvis and other equatorial island possessions grew 
as trans-Pacific aviation became a reality. In 1935, Jarvis, Howland 
and Baker Islands were colonized with Hawaiian high school graduates. 
This action was followed by the Presidential Order 13.5 in 1936 
re-establishing the claim to the islands as American territory (Bryan 
1974). Great Britain then relinquished its claims (Leff 1940). The 
islands have since been administrated by the Department of the Interior 
and are now part of the Hawaiian and Pacific Islands National Wildlife 
Refuge system directly administered by the U.S. Fish and Wildlife 
Service in Honolulu, Hawaii. 

Human occupation of Jarvis Island created changes in the simple 
ecosystem. Goats, rats, cats, mice and introduced plants like 
Abut i Ion indicum undoubtedly affected the native ecosystem (Bryan 
1974; Christophersen 1927; Judd 1960). In 1885, no birds were seen 
during a survey in October-November. One cat was observed "in 
possession of a house" (MacFarlane 1887). The Whippoorwill Expedition 
of 1924 found many mice and seabirds but no cats (Gregory, 1925). 



CATS ON JARVIS ISLAND 



By 1935, goats and cats had been extirpated but Polynesian rats 
( Rattus exulans ) were still abundant. "What we call 'field mice 1 
[small Polynesian rats] by the dozen crawl over the beds during the 
night and sometimes get caught between the blankets." (Bryan 1974). If 
cats had become established in 1885, rats would not have been so 
conspicuous. The Hawaiian colonists were required to note relevant 
biological observations (Bryan 1974). The presence of cats would not 
have escaped their notice. However, cats are mentioned in the 
correspondence surrounding the death of Karl Kalawai on October 1938. 
Kalawai's contemporaries were required to write their version of his 



death from appendicitis. In one letter, it was noted that the pet cat 
and dog were doing fine (Bryan 1974). It may be that the colonists 
brought cats, against orders, to prey on the disturbing number of rats. 
The cats may have escaped the confines of the camp and have become 
feral. King (1973) outlines the settlement of Jarvis and states that 
the settlers left cats when they abandoned the island. 

During the International Geophysical Year 1957-58, the island was 
manned as a research station with scientists from Scripps Institute of 
Oceanography (King 1973). One scientist reported killing several 
hundred cats during his stay (R. Clapp pers. comm. ). The Pacific Ocean 
Biological Survey Program (POBSP) visited Jarvis in 1964 and 1965 and 
they killed over 200 cats (King 1973). POBSP visits in 1967 and 1968 
located only 9 cats in 2 days (King 1973). A visit in 1973 by the U.S. 
Bureau of Sport Fisheries, now the U.S. Fish and Wildlife Service, 
sighted at least 14 cats (Kridler 1973). Giezentanner (1976) reported 
killing 12 cats and sighting 50 more on 2 surveys around the island in 
1976. In 1977, 102 cats were shot and 50 to 75 remained alive. In 
1978, 160 cats were shot, most of them kittens (Forsell 1978). These 
data suggest the carrying capacity is less than 200 cats. 

Surveys conducted opportunistically by the POBSP and during 
subsequent eradication efforts have identified and partially defined 
the extent of cat predation. Rats were formerly abundant on Jarvis but 
were probably extirpated by cats (King 1973). Mice ( Mus musculus ) 
persist in varying numbers. The POBSP noted the largest populations of 
birds over 4 years from 1963 to 1966 (Table 1). Three species of birds 
continue to breed in substantial numbers. Masked boobies ( Sula 
dactylatra ) and red-tailed tropicbirds ( Phaethon rubricauda ) are 
relatively large birds capable of defending themselves and their young. 
Sooty terns ( Sterna fuscata ) breed in very large numbers and inundate 
the cat population with more food than it can consume, then depart en 
masse. Nevertheless, the cats' predatory effects are significant. On 
nearby Starbuck Island, a cat population about the same size as that of 
Jarvis killed about 1000 birds a night (King 1973). Subsequent bird 
surveys have not been as comprehensive as the POBSP work which makes 
comparisons difficult. However, gross changes are apparent. In spite 
of annual and seasonal variation, the most obvious trend appeared to be 
a precipitious decline to the point of extirpation of the red-footed 
booby ( S. sula ) and the frigatebirds ( Fregata sp.). Only roosting 
birds were seen in 1982-83 though derelict nests were found. Small 
ground-nesting birds were absent. Surveys in November 1983 indicate 
that predation has ceased and may offer proof that cats have been 
eliminated. 



METHODS AND MATERIALS 



From June 14 to 27, 1982, a study of the ecology and behavior of 
the cats was undertaken by David Woodside of the U.S. Fish and Wildlife 



Service (FWS) and myself. Radio telemetry was used to determine home 
range and to test the efficacy of feline panleucopenia virus as a 
possible control agent. From June 27 to July 10, 1982, eradication 
techniques were implemented while additional biological observations 
were made. Attempts were made from October 28 to November 3, 1982, by 
Woodside, Steven Fairaizl (FWS) and Utimawa Bukaireiti of the Christmas 
Island Wildlife Conservation Unit (CIWCU) to complete the eradication 
work using recommendations garnered from the first trip. From March 3 
to 9, 1983, Woodside and Katino Teebaki (CIWCU) attempted to remove the 
final cat(s) from Jarvis Island. Cameron B. Kepler (FWS) searched for 
cats from November 6 to 10, 1983, without sighting any. His seabird 
observations indicate that cat predation has ceased and that cats may 
be absent from Jarvis Island. 



RADIO TELEMETRY 



Cats were captured using TOMAHAWK live traps set nightly with 
various baits. Traps were checked early each morning to avoid 
subjecting the animals to heat stress. When a cat was to be collared 
with a radio transmitter, it was removed from the trap by placing a 
burlap bag over one end. The trapdoor was opened and the cat was 
directed into the bag. After the bag was tied shut, it was weighed 
using a 5 kg PESOLA spring scale. The bag weight was subtracted to 
give the net weight of the cat. The bag was carefully opened and the 
cat's head positioned at the entrance of the bag. The head and neck 
were exposed while the body continued to be restrained. While the head 
was held in a gloved hand, the neck was fitted with a radio transmitter 
collar. The collar was measured and cut to the appropriate size 
allowing the cat adequate room to swallow. The ends of the collar were 
bolted and glued together. The exposed metal end of the collar was 
taped and covered with heat-shrink plastic tubing to prevent any 
moisture from interfering with radio transmission. 

Seven cats were fitted with radio transmitters of distinct 
frequencies for individual discrimination. AVM radio transmitters were 
selected for their range of frequency within the assigned U.S. Fish and 
Wildlife Service bands (164.467 to 164.709 megahertz) as well as their 
3-month battery life. The collar weighed about 25 g which is the 
maximum size a cat can carry without interfering with its ability to 
carry out normal behavior. The transmission could be received over a 
3 km distance or line of sight. Complimentary receivers (AVM) powered 
by 8 AA DURACELL batteries were used in conjunction with YAGI 3-tined 
antennaes compatible with the selected frequencies. 

Radio checks were conducted twice a week by scanning the occupied 
channels. Compass bearings were taken of the maximum signal strength 
determined by standard pattern sweeps of the antenna (Cochran and Lord 
1963). This single bearing technique sufficed when a general position 
was sought. Fixes were taken from 1 or 2 stations at known map 
locations and then plotted to determine a more exact position (Figure 



4). However, some sources of bias and sampling error in this method of 
triangulation were sufficient to justify the more time consuming 
process of homing (Springer 1979). Homing was necessary to locate the 
den sites of inoculated cats. This technique is to follow the 
direction of the maximum signal strength as it audibly increases. By 
intermittently increasing the interference (the 'squelch' knob) as the 
signal audibly increased, we were able to hear nuances in the strength 
of the signal and hence, determine it directionality (Cochran and Lord 
1963). 

The nature of the radio signal indicated the activity of the cat. 
A signal that varies in strength indicates an active animal that is 
moving its head. A steady signal indicates a recumbent animal. The 
surrounding environment affected the nature of the signal as well. 
Dense coral slabs interfered with the signal to present a 'muffled' 
sound. Likewise, transmission around metal objects gave a 'bounce' 
which confused directionality. 



FELINE PANLEUCOPENIA 



Five cats fitted with radio transmitters were given an oral dose 
of 0.5 cc feline panleucopenia virus (FPLV) and released. Two cats 
were collared and released unexposed to FPLV as control animals. An 
additional 26 cats were inoculated, 19 of these were marked with spray 
paint and released (Table 2). Data from bi-weekly fixes were used to 
determine the areas to which FPLV might spread as well as to define 
home range, den site affinity and movement. Significant movement 
hastened the spread of the contagion and justified inoculating fewer 
animals. 

Poole (1972) has shown that a modified live panleucopenia virus is 
stable at 37 C when stored in the medium in which it was produced for 
up to 5 months. Based on in vitro stability of the virus when stored 
at room temperature, we felt that the virus could retain virulence when 
stored without refrigeration at Jarvis Island. Since we were unable to 
hold cats in captivity to determine the dosage and potency of the 
virus, we used the radio-collared cats to provide these data. 



BAITS 



Bait attraction studies were conducted concurrently with live 
trapping to determine the most effective baits for this and future 
control efforts (Table 3). The first baits were chosen for their 
intrinsic attractiveness to cats. Several species of reef fish were 
initially tried. Gray mullet ( Mugil cephalus ), ' aholehole ( Kuhlia 
sandvicensis ) and man in i ( Acanthurus triostegus ) were used as 
baits for 80 trap-nights. The fish were halved and partially scaled to 
enhance their attractiveness to cats. 



8 

In order to determine if traps were affecting the attractiveness 
of baits, a series of bait trials was established. Each evening for 
one week, 2 fish baits were set at least 10 feet from each other at 9 
designated bait stations. Any evidence of visitation or fish 
consumption was recorded. Canned cat food (CALCAN; 'simmered supper', 
liver and fish) and canned sardines were also tested. A series of 6 
open cans were placed at bait stations away from the traps. Station 
areas were cleared to detect cat footprints. Tincture of catnip was 
used as a lure in spite of previous low success rates (van Aarde pers. 
comm.). Twenty drops of catnip were placed on a small cloth bag 
stuffed with grass ( Lepturus repens ) to aid fragrance dispersal. 
Finally, the attactiveness of freshly killed sooty terns which were cut 
in half and set at bait stations was determined. During the second 
trip, feline gland lure was used on steel and CONIBEAR 220 traps. This 
experimental lure was made from the testes and urine of male cats. 



POISON 



After 14 days had elapsed, chemical and mechanical control methods 
were initiated. Because the use of the effective predacide 1080 is 
banned on federal lands, 3 experimental compounds and 2 known toxicants 
were cage bioassayed in Hawaii in order to determine a suitable 
alternative (Fellows 1982). Based on these results, the compound 
N-(3-chloro-4-methylphenyl acetamide) or CAT was chosen for field 
testing. It was imperative that CAT hold no secondary poisoning 
potential for scavenging birds or invertebrates. A series of cage 
bioassay tests were conducted with 7 captive hermit crabs ( Coenobita 
perlitus ) . Fifty grams of dry cat food was mixed with seawater and 
90 mg of CAT . This yielded a toxicity of 0.18% Active Ingredient 
(Al). The crabs were held for 3 days in a shaded chicken wire cage and 
fed this mixture and water. The experiment was repeated with 3 groups 
of 6 crabs. The control group received 50 g of cat food and water. 
The second group received 50 g of moistened food with 90 mg (0.18% Al) 
of CAT . The third group received moistened food with 180 mg (0.25% 
Al) of CAT . All groups were held for 3 days and released. The 
control group was unmarked, the second group was spray-painted orange 
and the third was spray-painted blue. 

Approximately 50 g of sooty tern flesh rubbed with 90 mg of CAT 
constituted the LD or lethal dose required to kill 50% of the cats 
upon first ingestion. Thirty pieces of freshly killed sooty tern were 
smeared with 50 mg of CAT and placed in open trays in the quarry area 
on June 27. Signs of visitation were noted the following morning. On 
July 10, the day prior to our departure, 20 pieces of sooty tern bait 
were placed in and around the burrows of wedge-tailed shearwater 
( Puffinus pacificus ) which were also used by cats. 



TRAPPING 



In order to capture gun-shy cats, 2 other styles of traps were 
used. Steel or 'gin' traps designed to capture the cat by the foot 
were used in situations when the trap could be camouflaged. Lethal 
CONIBEAR 220 traps were set in den site entrances to trap cats as they 
entered or left. 

To determine the density of house mice on Jarvis Island, a line of 
snap-traps baited with peanut butter was set in the quarry area within 
a microhabitat of Lepturus repens and around the camp at the north 
coast. The 3 traps at camp were caged inside chicken wire to exclude 
hermit crabs. An additional 17 were set without wire cages since 
hermit crabs are largely absent from the quarry. 

Kepler (1984) attempted to census the house mouse population by 
counting mice on 5 transects covering an area of approximately 16,000 
ft . 



HUNTING 



Hunting has played a major role in the eradication of cats from 
numerous islands in New Zealand (Veitch 1980) and South Africa (van 
Aarde 1980). Its effectiveness is enhanced when used in conjunction 
with a battery of other control measures (Beck 1975). Night hunting 
began on 27 June 1982 and continued until 10 July. Since cats are 
primarily nocturnal, headlamps powered with 2 D cell batteries were 
used to illuminate the horizon and create 'eyeshine,' a reflective 
response from the inner layer of the cornea of nocturnal animals. This 
is visible as a blue or orange glow from about 50 m away on dark 
nights. Ambient moonlight diluted the reflective response. On the 
second visit, which coincided with a full moon, more powerful spot 
lights were used. Night hunting was done with a Q-BEAM and SL-20 
rechargable hand-held spotlights. The SL-20's were taped to the 
barrels of the guns to facilitate sightings. A REMINGTON 12 gauge 
shotgun, single barrel with a modified choke using No. 2-3 shot for 
ample pattern spread and stopping power, was used primarily at night. 
Cats were shot during the day with a 0.22 mm calibre rifle fitted with 
a 4 by 40 telescopic sight using long hollow point ammunition. 
Post-mortem data collected from each cat include weight, sex, color, 
reproductive condition (in females) and stomach contents (Table 4). 



BIRD CENSUSES 



A census of nesting birds was made to determine what effects the 
removal of predators would have on population numbers. Comparison with 
earlier POBSP data allowed population trends to be identified (Table 



10 

1). Estimates of nesting sooty terns were made by measuring four 10 
meter square plots and counting all the enclosed eggs. Simple counts 
of masked and brown boobies ( S. leucogaster ) , wedge-tailed 
shearwaters and red-tailed tropicbirds were also made in 1982. Other 
non-breeding birds were censused as they arrived. The population 
estimates represent maximum numbers. Kepler (1984) censused birds in 
November 1983 using a series of transects. His work will allow 
repeated comparisons on future expeditions. 



RESULTS AND DISCUSSION 
RADIO TELEMETRY 



The results of the telemetry study indicate that the 
radio-collared cats remain in specific den sites during the hottest 
part of the day and become active at dusk (Figure 4). The principal 
feeding area on Jarvis Island was located along the south shore away 
from any known den sites. Panaman (1981) reports that within the home 
range, cats attempt to cover droppings but outside the home range, 
droppings remain exposed. Several high-density areas of cat droppings 
were found in the middle of the island well away from any suitable den 
site cover. It may be that the north shore cats cross the island to 
feed and while in transit, pause to mark these sites. No dropping 
sites were found in known home ranges of collared cats. 

The home range was determined by plotting the fixes taken by 
homing and triangulation. Some positions taken by triangulation may be 
in slight error. Also some signals were not received during fixes. 
The female cats (nos. 3, 4, 7, 9) were consistently tracked to specific 
dens. Cat No. 3 used the periphery of the guano quarry as a home range 
(Figure 4). Cat No. 4, a lactating female with a small kitten, used 
the southern portion of the quarry. This area had dense Lepturus and 
loose ground for cover. Cat Nos. 7 and 9 inhabited the coral slabs of 
the north coast. There appeared to be considerable movement along this 
coast. No. 7 was also recorded in the quarry by a questionable fix. 
The home range area surrounding the den site is larger for males than 
for females (Macdonald and Apps 1978). The male cats (Nos. 6, 8, 11) 
were located in various den sites. Cats 6 and 8 were once recorded 
sharing the same den at the same time in the quarry area. Cat 8 was 
recorded mostly from the north east coast but was killed in the quarry 
accompanying cat No. 4. 

It may be that the lack of suitable dens forced cats to share. 
One den in the quarry was occupied by at least 2 males, 2 females and 
one large kitten during the observation period. Although no cats were 
recorded near the southern sooty tern colony, it is clear from stomach 
analysis that the cats move freely across the island to feed on terns. 



11 



Veitch (1980) found cats on Little Barrier Island had a variable home 
range depending on the proximity of the feeding area. 



in:: i ;.-2Z-Z~:zz;z1~za 



We felt the spread of the contagion FPLV was likely because cats 
seeaed to co-habit sites. Transmission of FPLV usually occurs by 
direct contact between the infected and the susceptible cats via 
saliva, feces, urine and fleas (Kahn 1978). In fully suspectible cats, 
i.e., those without active immunity, termination of the disease, either 
by death or by recovery is from one to 10 days (Veitch 1982). Death 
say occur at any stage after the rise in body temperature, however, it 
usually occurs after 2 to 4 days of manifest illness. The virus 
creates severe hemoto logical changes by destroying the blood-formi-z 
tissues. There is a gradual fall in the white blood cells, followed by 
dehydration. Recovery is characterized by a rise in white blood cells 
and the appearance of antibodies. It takes the animal several weeks to 
regain normal body weight but it will have acquired active immunity for 
up to 4 years (Gaud and Hallauer 1976; Veitch 1982). Feline 
panleucopenia is present in feral cat populations of most large areas. 
The disease tends to gradually build up to epidemic proportions and 
spread to all susceptible individuals before it dies out. Islands are 
usually too small to harbor reservoirs of the virus, so once it has 
passed through a population, it will die out. We believed the cats on 
Jarvis may have been exposed to the virus many years ago but were again 
susceptible. 

The virus appears to have had a significant effect on the 
population. Cat No. 3, a female, died within 10 days of receivi-g a~ 
oral dose of FPLV. She exhibited the clinical symptoms of mucal 
discharge about the eyes and nose. She was located via telemetry at the 
periphery of the quarry appearing sluggish and listless. The next day 
she was found dead in the open grass. The remaining 4 radio-collar ed 
infected cats were allowed to live up to 18 days before it was 
necessary to kill tze~. No further expression of FPLV was noted. 
During the hunting phase, 10 out of the 19 marked and infected cats 
were shot. All appeared healthy with excess fat stores in the 
peritoneal cavity and displayed glossy coats and well— developed teeth. 
However, 9 of the 19 were not resighted. It may be that these cats 
died from the disease and went unnoticed. It is also possible that 
these cats were among those shot at night and not recognized because 
the marks rubbed off. The possibility that these cats represent 
additional FPLV mortality must be considered in light of the data 
presented by Scott, et al. (1970). He reports that about 50% of the 
challenged cats will expire. Van Aarde (pers comm) has used FPLV to 
reduce the cat population of Marion Island by 54% within 2 years of 
initial exposure. His most recent survey (May 1980) indicated a 
further 65% decrease with no indication of immunity build-up. At Jarvis 
Island, we felt that mortality would be above 50% since the xeric 
climate would hasten dehydration. The mean daytime temperature of 



12 

95°F may have affected the virulence of the virus in spite of Poole's 
(1972) suggestion that the virus could survive at 99 C. At best, we 
had a 41% mortality of marked cats (Table 2). 

Veitch (1980) concluded that FPLV did not work well enough to 
justify the trapping effort needed to inoculate at least 5% of the 
population to spread the contagion. In areas with dense cover, it is 
very difficult to determine the extent of mortality necessary to 
justify its continued use in lieu of more conventional control methods, 



BAITING AND TRAPPING 



Baits are used either to lure cats into traps or to carry poison 
(Veitch 1982). Fresh fish is a readily available bait on islands so we 
began to trap cats using fresh reef fish (Table 3). In the 80 traps 
that were set, 11 cats were captured (14% success rate). Concurrently, 
18 pieces of fish were set outside of traps to see if cats avoided the 
traps. Only 2 pieces (11%) appeared to have been visited by cats. 
Canned cat food and sardines were tried as bait in 12 traps but no cats 
were captured. Even at 18 bait stations outside of traps, there was no 
evidence of cat interest. Tincture of catnip was used in 4 traps also 
without success. 

Sooty terns appear to be the most attractive bait judging from 
trapping results. Before hunting began, 64 traps were set with terns 
and 25 cats were captured (39%). After hunting, 40 traps were set and 
4 cats were captured (10%). The relatively high success rate of 39% 
led us to choose terns as the main bait for trapping and poisoning 
efforts. Overnight, all baits became infested with Oedemerid beetles 
( Ananca bicolor ) which greatly reduced their attractiveness. 
However, day-old sooty tern pieces still held some attraction. This 
was an unexpected lure since felids do not readily accept carrion as 
bait (Beck 1975). Two cats were captured with this bait and several 
others showed interest by pulling the feathers which protruded from the 
cage until the meat was in contact with the wire mesh. Some fresh 
baits were partially consumed in this manner. This learned behavior 
may indicate trap avoidance. Only 2 marked inoculated cats were 
recaptured, although unmarked inoculated cats may have been. Using all 
combined baits, 200 traps were set and 40 cats were captured (20%). It 
is instructive to compare the trapping effort at Little Barrier Island, 
a heavily forested and highly eroded island, with that at Jarvis. In 
1977, 2637 traps were set and 26 cats were caught. In 1978, 37,332 
traps were set and 73 cats were caught. In 1980, the last year of 
trapping, 32,615 traps were set and only 5 cats were caught (Veitch 
1982). 

Trapping success is a function of population size and the 
experience of the trappers. As animals become more scarce, the success 
in trapping declines at roughly an exponential rate. In order to trap 
experienced animals, we used 2 other styles of traps, leg-holds and 



13 

CONIBEARS. These traps would have had a higher success rate if used 
earlier in the eradication campaign. Since these traps were not 
baited, they were placed in areas that animals frequent, like denning 
sites or runways. Leg-hold traps were placed in entrances and along 
runways blocked with fencing which detoured the cat into the traps. 
During 49 trap-nights, one cat was caught (2%). CONIBEAR 220 traps are 
even more site specific since they must be supported externally. Seven 
traps were placed in the entrance to dens. Two cats were killed. Two 
red-tailed tropicbirds were caught and killed while exploring potential 
nest sites. These were the only cases of non-target vertebrate 
mortality. 

During the October-November trip, feline gland lure was used in 
addition to various baits such as booby meat and cooked fish. In spite 
of the lure's previous success in mainland situations, no cats were 
caught. This is probably a result of very low population densities. 

In attempting to estimate rodent population density, 46 snap-traps 
were set in the quarry. No mice were caught. In traps without a 
chicken wire enclosure, 4 land crabs were caught. Three traps were set 
at camp where mice were previously seen. Two were caught. 



POISONING 



Since hermit crabs are potential subjects for secondary poisoning 
by scavenging, a series of toxicity tests were initiated on 18 June 
with 7 crabs. The crabs received cat food mixed with 0.18% AI CAT . 
They were not observed eating the bait during the 3 day trial. They 
appeared listless and hung upside-down in the cage after repeated 
escape attempted failed. They were released and the food was exposed 
to free-roaming crabs who quickly consumed it. It appeared that the 
listless behavior was a response to captivity. 

This experiment was repeated with 3 groups of 6 crabs. A portion 
of the poisoned bait was consumed by 2 groups. The crabs again 
appeared listless. One individual shed its shell and escaped through 
the wire mesh. The spray-painted crabs which had consumed bait with 
0.18% and 0.25% AI CAT were released and subsequently resighted one 
week later along the beach. It appeared that these concentrations of 
CAT were not lethal to hermit crabs. These data are essential if 
aerial broadcasting of poisoned bait is considered as a future control 
method. 

Thirty 50 g pieces of sooty tern were poisoned with 0.18% CAT 
(Table 3). The following morning, 15 pieces (50%) remained untouched. 
Five pieces (16%) were moved but not eaten. Nine pieces (30%) were 
partially eaten and one piece was wholly removed. Two days later, this 
test was repeated with 28 pieces. Seventeen (60%) were untouched. Two 
(7%) were moved but not eaten. Two were partially eaten and 7 (25%) 
were removed. The relatively high rate of consumption (33%) approaches 



14 

the rate of trapping success using sooty terns. In addition, 28 pieces 
of poisoned bait were placed in and around shearwater burrows occupied 
by cats. The effect of the baits is unknown. No carcasses of poisoned 
cats were found. Veitch (1980) reports similar findings. At least 
26,850 pieces of bait were placed on Little Barrier Island, but only 4 
carcasses were found. 

Fellows (1982) determined in test animals that the mortality rate 
was about 75% in cats that consumed at least 13 mg of CAT per kg of 
body weight. Assuming that only half of the 50 g bait (0.25% AI) was 
consumed, it would have delivered a lethal dose to the heaviest cat on 
Jarvis Island (Table 4). Death from CAT is due to renal failure 
(Palmore 1978). Like FPLV, it was hoped that the xeric climate might 
increase mortality. During the study, rainfall was sporadic, yet 
sufficient quantities collected in the shells of the giant clam 
( Tridacna maxima ) to provide a constant supply. 



HUNTING 



The number of cats shot during the 1982 hunting period is plotted 
in Figure 5. The number of cats shot per day is plotted along the Y 
axis with the number of hours hunted per day. The number of hunting 
days is plotted along the X axis. The obvious trend is initially high 
mortality with a quick drop-off as hunting progresses. The number of 
hours hunted per day is the man-hour effort. As targets become fewer, 
man-hours to hit those few targets increases. The first 7 days of 
hunting yielded 1.97 cats per man-hour of hunting. The yield for the 
second week was only 0.19 cats per man-hour. The total yield for 110 
man-hours of hunting was 105 cats or about one cat per hour of hunting 
from a population of about 44 cats per km . 

Past eradication efforts on Jarvis Island have been brief though 
targets numerous. Thus a high rate of depletion was obtained by 
Forsell (1977, 1978). He shot 5 cats per man-hour in 1977 for a total 
of 102. He estimated that 50 to 75 cats remained alive. In 1978, he 
and a group of U.S. Coast Guardsmen shot 4 cats per man-hour to reach a 
total of 160. With the same manpower on Howland Island, the hunting 
kill was only 0.8 cats per man-hour in 1977 and 0.14 in 1978. The 
latter rate approximates that in our attempt to shoot the last Jarvis 
cats. On the October 1982 trip, Woodside spent over 100 hours to shoot 
two cats. In March 1983, over 100 hours were hunted without success. 
At least one cat was sighted (Woodside pers. coram.). In June 1983, an 
experienced hunter reported seeing no cats during 2 days of hunting 
(Austin pers. comm.). Kepler (1984) hunted for 35 hours over 4 days 
without seeing any sign of cats, i.e., eyeshine or predated birds. 

Hunting on the Kerguelen Archipelago halved the maximum lifespan 
and lowered the population age as well as caused a disequilibrium in 



15 

the sex ratio. Hunting reduced the geographical range of the cat 
population (Pascal 1980). On Marion Island, van Aarde (pers. comm.) 
reported; "Under sub-Antarctic conditions with population density of 
approximately 10 adult cats per km , a success rate of 2.5 hours per 
cat (0.4 cats per hour) was achieved. This efficiency decreased 
roughly exponentially with a decrease in population density." Both 
Marion and Jarvis Islands are relatively clear of vegetative cover so 
hunting can be used effectively. On the well-vegetated Little Barrier 
Island even hunting with dogs was futile. 



BIOLOGICAL POPULATION CHARACTERISTICS 



Hunting provided the opportunity to examine the biological 
characteristics of the Jarvis Island cat population as a whole (Tables 
4 and 5). The color and sex of 108 cats were recorded. Black females 
were the most common phenotype (33%) followed by black males (25.5%). 
Black cats composed 58.5% and tabby cats composed 31.5% of the 
population. While live-trapping, we noted that black cats caught in 
traps and exposed to the morning sun appeared listless and frothy at 
the mouth, but lighter tabby cats showed no ill effects. The black 
cats possibly experience more heat stress from high solar radiation 
than the lighter tabby cats. Van Aarde (1980) hypothesized that dark 
coat color may have some advantage in the sub-Antarctic and that a 
strong founder effect is indicated by the absence of piebald spotting. 
On Jarvis, only 2.1% of the cats were piebald. On the Kerguelen 
Archipelago, the feral cat has kept the principal dark color 
characteristics of the domestic cat for over 20 years (Derenne 1976). 
Dark cats may be more successful nocturnal hunters than lighter ones. 

Relatively small sample sizes prevent any conclusions from being 
drawn but some interesting trends are apparent. Overall, the sex ratio 
is roughly equivalent; 52% females, 48% males. This difference is not 
significant (P<0.5). However, the gray cats show a highly significant 
(P>0.02) sexual bias to males (9:1). Piebald cats were the least 
common phenotype represented by 2 males and one female. The other 
non-gray cats were at least 95% black. 

The weights of 42 adult cats fell within the normal range for the 
common domestic cat (Scott 1972). Tabby males were on the average the 
heaviest but the heaviest individual was a black male. The weights of 
both sexes were heavier than those reported from Raoul and Little 
Barrier Islands in New Zealand and lighter than those from Herekopare 
and the sub-Antarctic Macquarie Island (Veitch 1982, Jones 1977). 



DIET 



Of the 54 cat stomachs examined, 32 (59%) contained flesh and 
feathers of sooty tern adults and embryos as well as eggs. A subcolony 



16 

of terns near the main colony was heavily predated. Small teeth marks 
on eggs indicated that cats fed on them. The colony was later deserted 
en masse. Although terns are known to desert colonies, especially in 
vulnerable peripherial areas, it would appear that predation is an 
added pressure to desert (Ashmole 1963). Stonehouse (1962) found that 
cats on Ascension Island rarely ate sooty tern eggs however there aie 
records of cats eating eggs of grey-faced petrels on Kerguelen Island 
and dominican gulls ( Larus dominicanus ) on Dassen Island, Southest 
Africa (Atkinson pers. comm.). Eggs may be a learned food source 
selected only by a few cats in some colonies. 

Analysis of stomach contents showed that twenty one stomachs ( 38% ) 
were empty. Since sooty terns settle on the ground after dark, the 
absence of food in these cats could indicate that night hunting had not 
yet begun. Also, during our hunting phase, the moon was full. 
Presumably, terns would be harder to catch on brightly lit nights and 
so the cats would be less successful. Since sooty terns are the 
primary food source, it is reasonable to assume that they limit the cat 
population. Stonehouse (1962) suggested that cats on Ascension Island 
are also limited by the nature of the tern breeding cycle. There is no 
shortage of food when terns arrive. When the breeding cycle is 
complete and they leave, cats are forced to survive on less easily 
obtained food. 

One red-tailed tropicbird chick was identified in a cat stomach. 
A deserted masked booby chick was observed being stalked by a cat. It 
was later missing. The remaining 3% of the stomachs contained parts of 
crickets and cockroaches. Fitzpatrick (1979) found many species of 
invertebrates in cats which indicated a seasonal dependence on this 
food supply. One gecko was also identified in the remains. In spite 
of the abundance of reef fish, only one cat had fish in its stomach. 
These were probably prey items from a sooty tern which itself had 
subsequently been consumed. It may be that the abundant hermit and 
ghost crabs effectively compete for the carrion of the beach. 

The last cat shot during the June/July trip had one mouse in its 
stomach. Several months later in October, one of the 2 cats shot had 5 
mice in its stomach. During the March 1983 trip, mice were reported as 
very common in contrast to earlier trips. This increase in mice is 
almost certainly related to the decreased predation pressure. In 
November 1983, mice were conspicuous and possibly undergoing a 
population crash. A rough estimate of 36,000 mice on Jarvis represents 
an order of magnitude figure (Kepler 1984). During the Whippoorwill 
Expedition of 1924, mice were abundant (Gregory 1925). In 1935, mice 
were still abundant (Bryan 1974). In quoting the journals of the 
colonists, Bryan inserted in parenthesis that mice were Polynesian 
rats; "What we call 'field mice' [small Polynesian rats] by the dozen 
crawl . . ." (Bryan 1974). If this were true, then the introduced cats 
eliminated the rats but not the mice. However, it may be that the mice 
eliminated the rats since evidence from Stewart Island, New Zealand, 
suggests that mice are able to exclude ecologically Polynesian rats 
from a grassland biome (Taylor 1975). In other tropical regions, 



17 



Polynesian rats do co-exist with mice but Jarvis Island may be such a 
simple ecosystem that this is impossible (Storer 1962; Tomich 1970). 



FECUNDITY 



During the hunting phase, 5 kittens were shot. This represents a 
relatively low recruitment rate which may be influenced by the lack of 
a steady food supply prior to the arrival of sooty terns. During late 
May 1982, terns began to arrive. Their arrival may have stimulated 
the onset of oestrous as evidenced by the apparent increase in 
pregnancies and the number of kittens in-utero. Of the 26 females 
examined, 8 were pregnant. The average number of embryos was 3. The 
survival rate of kittens is unknown but at least 24 could have been 
born. 



BIRD POPULATIONS 



Four 10 m by 10 m quadrats within the sooty tern colony were 
censused. The mean density of eggs was 37 per 10 m . We then 
measured the roughly linear colony to determine the area and multiplied 
that area by the mean egg density to determine that 210,000 eggs or 
444,000 nesting birds were present. We estimated that an additional 
third of the colony were nonbreeders. Near the western and eastern 
beaches were 2 more colonies (Figure 4). Approximately 500,000 
additional birds were present bringing the estimated total population 
of sooty tern to over 1 million birds. 

Stonehouse (1962) considers the predatory effects of even 100 cats 
to be considerable. If the cat population is about 120 on Jarvis 
Island, and each cat eats a bird a day for about 200 days; the average 
period a colony might be established, then the annual cumulative 
predation could approach 25,000 birds per year or about 2.5% of the 
total sooty tern population. 

The masked booby colony is one of the largest in the Central 
Pacific Ocean. King (1973) estimated that 9000 individuals were 
present. Our estimates of breeding birds agree but are slightly less 
for nonbreeding birds (Table 1). This species is loud and aggressive 
and apparently suffers little cat predation. Likewise, the red-tailed 
tropicbird is well-defended. King (1973) recorded "large populations of 
both frigatebirds and all three boobies. . . ." In 1982, about 1550 
lesser frigatebirds and 550 red-footed boobies roosted in the center of 
the island at night. These birds were not breeding on Jarvis in 1982 
and may be considered victims of cat predation. One partially consumed 
booby was found in the quarry well away from the roost site. In 
addition, less than 50 brown boobies and 10 wedge-tailed shearwaters 
breed on Jarvis Island. Other petrels, shearwaters and small terns are 
absent. A wing of a white-throated storm-petrel ( Nesofregata 



18 

albigularis ) was found on Jarvis Island indicating that this species 
may still visit and be a potential colonist. 

Kepler (1984) found increased numbers of birds present during his 
November 1983 surveys conducted at approximately the same time as in 
the previous year (Table 1). He found 4 lesser frigatebird colonies 
with chicks. The largest colony (286 pairs) was only 50 m east of the 
guano quarry where cats were common in 1982. Red-footed boobies were 
breeding in small numbers. Kepler found 2 colonies, one with 15 nests, 
the other with 7 nests. Over 800 roosting birds were counted at night. 
Four separate sooty tern colonies were found out of syncrony with each 
other. Kepler attributes this to the species recovery from the effects 
of the El Nino that began in August 1982 (Firing et al. 1983). 



CONCLUSIONS 



Jarvis Island is considered to be of outstanding importance for 
the abundance of its wildlife especially breeding seabirds. The 
elimination of feral cats will probably make Jarvis Island one of the 
largest seabird colonies in the central Pacific Ocean (King 1973). 
Currently, the island is administered as part of the U.S. National 
Wildlife Refuge System by the U.S. Fish and Wildlife Service. 
Beginning in 1973, the Service attempted to eradicate the Jarvis cats 
by using sporadic control measures. In 1981, a systemic analysis of 
available control options and a survey of the results from other cat 
eradication work was conducted. As a result of this preparation, we 
investigated the use of feline panleucopenia virus as a control agent. 
Research in other insular situations suggest this technique could be 
helpful in reducing a susceptible population by at least 50%. By 
determining the home ranges and movements of cats using radio 
telemetry, we were able gauge the potential spread of the virus through 
the population. Only one radio-collared cat was killed by the virus. 
Additional deaths may have occurred to marked cats that were not 
subsequently recovered. If so, then the total mortality via FPLV was, 
at best, 41% of the inoculated cats. Nevertheless, we judged this 
technique to be relatively ineffective in lieu of other control 
measures which allowed full accountability of mortality. 
Accountability is essential, especially in the case of Jarvis Island, 
when visits are infrequent and of short duration. 

Poisoning was another indeterminate technique particularly because 
experimental compounds were used. The current controversy surrounding 
the use of compound 1080 on federal lands prevented us from obtaining 
this effective toxin. However, bioassays conducted in Hawaii and on 
Jarvis Island have indicated that CAT is an effective control agent 
for cats and does not hold any secondary toxic effects for scavenging 
invertebrates. Aerial baiting could be possible with this toxin. 



19 

The combination of hunting and trapping proved most useful in 
removing the majority of the cats from the island. However, the 
success of these methods is limited by the amount of manpower 
available. As targets become fewer, the effort must correspondingly 
increase if the last cats are to be shot or trapped. This effort can 
become very costly, time consuming and frustrating especially if time 
is limited. On Jarvis Island, the last 250 man-hours of hunting 
removed only 2 cats. 

Trapping efficacy is also affected by the population density of 
animals. After one week of hunting, the capture rate decreased from 
39% to 10% while using terns as bait. A variety of baits were tested 
to determine the most attractive. Familiar foods, i.e., terns and fish 
(to a lesser degree), attracted all the cats that were trapped. The 
use of feline gland lure was ineffective but might succeed when baits 
fail during periods of abundant prey. Steel leg-hold and CONIBEAR 220 
traps were useful and could have caught more cats if used earlier in 
the eradication effort. They could be effective in removing the last 
cats provided ample time was available. It is essential to use all 
available methods over a reasonable period of time to provide a broad 
front of eradication techniques to remove the last wary cats. 

The ability to account for the carcasses of cats afforded an 
opportunity to survey the phenotypic expression of an entire 
population. The observation that the majority of the population is 
black suggests some adaptation to the environment. In spite of the 
black cats' susceptibility to heat stress during the day, they may be 
less conspicuous than the other phenotype when hunting at night. All 
cats examined appeared healthy with sleek fur and adequate fat 
deposits. The weights of the cats fell within the normal range for the 
domestic house cats, and in fact were heavier than temperate climate 
cats from New Zealand. 

Ecologically, Jarvis Island is severely changed. The miners who 
removed 300 thousand tons of guano initially altered the simple 
ecosystem with introduced goats, rats and mice. Yet seabirds were able 
to continue to utilized the island as a nesting ground until the 
introduction of cats. Cats allegedly extirpated the smaller nesting 
species which occur on similar tropical islands without predators. It 
was the purpose of this eradication effort to remove the cats in the 
hope that these species of seabirds, which are threatened by predators 
on many other islets through the Pacific, would return to colonize 
Jarvis Island. Future surveys will be needed to identify the extent of 
recolonization and to monitor the status of introduced species. 

In attempting to rid the island of cats, we tried various 
established and novel techniques. In 1982, we removed about 120 cats 
but were not successful in complete eradication in spite of 6 weeks of 
effort involving 5 people. However, it appears that very few, if any, 
cats remain. The first major nesting effort by lesser frigatebirds in 
2 years suggested that the cat population has either been eradicated or 
is very low and probably not breeding. If the population contains one 



20 

pregnant female, the population could rebuild its numbers in a brief 
period. We are not safe in assuming complete eradication until 4 or 5 
years have passed with no cat signs. The threat of future cat or rat 
introductions is always present so long as ships pass by the island. 
The remoteness of Jarvis Island makes protection efforts very 
difficult. 

Habitat rehabilitation is the responsibility of agencies in charge 
of administering disturbed lands. In the end, it remains the duty of 
the U.S. Fish and Wildlife Service to monitor the effects of 
eradication on Jarvis Island. To return disturbed ecosystems to a more 
natural state is a difficult task. Yet, every effort must be made to 
erase the deleterious effects that animals introduced by man have 
wrought. It must be stressed that man was the initial source of the 
introduced cats on Jarvis Island. Thus, it is our responsibility to 
remove them so the island can once again become a predator-free colony 
for many species of tropical seabirds. 



ACKNOWLEDGEMENTS 



I would like to thank the U.S. Fish & Wildlife Service for 
logistical and financial assistance in reaching Jarvis Island. 
Transportation was kindly provided by the R/V MACHAIS, Captain Bill 
Austin and crew under contract to the Hawaiian Institute of Geophysics. 
Particular thanks are due to Drs. Eric Firing and Dean Roemmich for 
allowing us to join their charter. The University of Hawaii Women's 
Campus Club generously supported this publication with a grant. 
Drs. Sheila Conant , Cameron Kepler, John Street and Lyndon Wester 
critically appraised the manuscript. Finally, I am especially indebted 
to Mr. David Woodside for his expertise and valuable help on Jarvis 
Island . 



21 
LITERATURE CITED 



Aarde, van R. J. 1980. The diet and feeding behavior of feral cats 
( Felis catus ) at Marion Island. S.-Afr. Tydskr. Natuurnau . 
10:3-4. 



1980. Gene frequencies in feral cats on Marion 



Island, South Africa. J. Hered . 5:366-68. 

Ashmole, N. P. 1963. Biology of the wide awake or sooty tern ( Sterna 
fuscata ) on Ascension Island. Ibis . 103b (3):297-362. 

Beck, J. (ed.). 1975. Use and development of sodium monof luoroacetate 
(1080) as a predacide. ASTM . 27-33. 

Bourne, W. R. P. 1975. Mammals on islands. New Scientist . 
165:422-425. 

Bryan, E. H., Jr. 1942. American Polynesian: Coral Islands of the 
Central Pacific . Tongg Publishing Co. Honolulu, HI 253 pp. 

. 1974. Panala'au memoirs . Bishop Mus. Press. 



Honolulu, HI. 249 pp. 

Byrne, R. 1980. Man and the variable vulnerability of island life — A 
study of recent vegetation change in the Bahamas. Atoll Research 
Bull . No. 240. 200 pp. 

Christopher sen, E. 1927. Vegetation of Pacific equatorial islands . 
Bishop Mus. Bull. No. 44. Whippoorwill Expedition #2. 71 pp. 

Cochran, W. W., and R. D. Lord Jr. 1963. A radio-tracking system for 
wild animals. J. Wildl. Mgmt . 27:9-24. 

Derenne, P. 1976. Notes on the biology of feral cats from the 
Kerguelen Archipelago. Mammalia . 40(4) :531-95. 

Egler, F. E. 1942. Indigene verses alien in the development of arid 
Hawaiian vegetation. Ecology . 23:14-23. 

Fellows, D. 1982. Candidate cat control agent bioassay . U.S. Fish 
and Wildlife Damage Research Station, Hilo, HI 16 pp. 

Firing, E., R. Lukas, J. Sadler and K. Wyrtki. 1983. Equatorial 
undercurrent disappears during 1982-1983 El Nino. Science 
222(4628). 

Fitzgerald, B. M., and B. J. Karl. 1979. Foods of feral house cats in 
the forests of the Orongorongo Valley, Wellington. N.Z. J. 
Zool. 6:107-126. 



22 

Forsell, D. 1977. Expedition report 1977 . U.S. Fish and Wildlife 
Service Refuge Division, Honolulu, HI. 

. 1978. Expedition report 1978 . U.S. Fish and 



Wildlife Service Refuge Division, Honolulu, HI. 

Gaud, S., and C. Hallauer. 1976. The parvovirus. Virology 
Monographs . Springer-Verlag. Wien, New York. 

Gregory, H. E. 1925. Report to the director for 1924 . Bishop Mus. 
Bull. #21. Honolulu, HI. 55 pp. 

Giezentanner , J. B. 1976. Expedition report: Jarvis Island . U.S. 
Fish and Wildlife Service Refuge Division, Honolulu, HI. 

Hague, J. D. 1862. Phosphate islands of the Pacific Ocean. Amer. J. 
Science . 34:224-243. 

Hutchinson, G. E. 1950. The biochemistry of vertebrate excretion . 
Bull. Amer. Mus. Nat. Hist. 96:255-276. 

Jarvis, P. T. 1979. The ecology of plants and animal introductions. 
Prog. Phvs. Geog . 3:187-214. 

Jewell, H. G. Jr. 1961. Preliminary report on the marine mollusca of 
the Line Islands. Haw. Shall News . 9(3):5, (4):l-7. 

Jones, E. 1977. Ecology of the feral cat ( Felis catus ) on Macquarie 
Island. Aust. Wildl. Res . 4:249-262. 

Judd , A. F. 1960. The guano islands: diary of A. Francis Judd. 
Voyage to Jarvis Island. 1858 . Honolulu, HI. 105 pp. 

Kahn, D. E. 1978. Pathogenesis of feline panleucopenia. J. Amer. 
Vet. Med. Assoc . 5(2) :628-630. 

Kepler, C. B. 1984. . Tarvjs Island report . U.S. Fish & Wildlife 
Service Research Division. Kula, Maui, HI. 

King, W. B. 1973. Conservation status of the birds of Central Pacific 
Islands. Wilson Bull . 85(1) :89-103. 

Kridler, E. 1973. Ascertainment report . U.S. Fish and Wildlife 
Service Refuge Division. Honolulu, HI. 

Leff, D. N. 1040. Uncle Sam's Pacific islets . Stanford Univ. 
Press. Stanford, CA. 71 pp. 

MacFarlane, J. R. H. 1887. Notes of birds in the western Pacific, 
made in H. M. S. 'Constance,' 1883-85. Ibis . 201. 



23 

Macdonald, D. W., and P. J. Apps. 1978. The social behavior of a 
group of semi-dependent farm cats, ( Felis catus ): a progress 
report. Carnivore Genetics Newsletter . 3:256-268. 

Mueller-Dombois, D., and G. 0. Spatz. 1972. The influence of feral 
goats on the lowland vegetation in Hawaii Volcano National Park . 
US/IBP Island Ecosystems. 1RP Tech. Rept. 13. 46 pp. 

Murphy, R.C. 1964. Discussion: conservation. In: Arctic-Antarctic 
Biology . R. Carrisk (ed.) Herman, Paris. 

Pascal, M. 1980. Structure and population dynamics of feral cats in 
the Kerguelen Archipelago. Mammalia . 2:161-82. 

Palmore, W. P. 1978. Diagnosis of toxic acute renal failure in cats. 
Fla. Vet. Journal . 8:14-15, 36-37. 

Panaman, R. 1981. Behavior and ecology of free-ranging female farm 
cats ( Felis catus L.). Zeitschrift fur Tierpsycholgie . 
56:59-73. 

Poole, G. M. 1972. Stability of a modified live panleucopenia virus 
stored in liquid phase. Applied Microbiology . 24(4) :663-64. 

Simberloff, D. S. 1970. Taxonomic diversity of island biotas. 
Evolution . 24:23-47. 

Scott, F. W., C. K. Cziza, and J. H. Gillespie. 1970. Feline viruses 
IV. isolation and characterization of feline panleucopenia virus 
in tissue culture and comparison of cytopathogenicity with feline 
pirornavirus, herpesvirus and pathoreovirus, Cornell Vet . 
60:165-183. 

Scott, P. P. 1972. The cat. In: The U.S.A.W. Handbook on care and 
management of laboratory animals (4th ed.). Churchill 
Livingston: Edinburgh and London. 

Springer, J. S. 1979. Some sources of bias and sampling error in 
radiotriangulation. J. Wildl. Mgmt . 43(1) :926-935. 

Stonehouse, B. 1962. Ascension island and the British ornithologists' 
union centenary expedition 1957-59. Ibis 103 B: 107-133. 

Storer, T. I. (ed.). 1962. Pacific island rate ecology . Bishop 
Mus. Bull. No. 225. 274 pp. 

Taylor, R. C. 1973. An atlas of Pacific island rainfall . Data 
Rept. No. 25. HIG-73-9. 

Taylor, R. H. 1975. What limits kiore ( Rattus exulans ) distribution 
in New Zealand. N.Z. J. Zool. 4:473-477. 



24 

Tomich, P. Q. 1970. Movement patterns of field rodents in Hawaii. 
Pac. Sci . 24:195-234. 

Veitch, C. R. 1980. The eradication of cats from Little Barrier 
Island . New Zealand Wildlife Service, Department of Internal 
Affairs, Wellington, N. Z. 

. 1982. Methods of eradicating feral cats from 

offshore islands in New Zealand . Proc. I.C.B.P. Cambridge. 

Vitousek, M. J., B. Kilonsky, and W. G. Leslie. 1980. Meteorlogical 
observations in the Line Islands, 1972-1980 . Data Rept. No. 38. 
HIG 80-7. 74 pp. 



25 



w 
m 
2 





















o 
























o 
























o 
















































<y* 






















Ol 


o 








CO 


• 


O 










CO 


CN 








00 


> 


CN 








o 


CO 


-^ 








CTi 


o 


r^ 








■■■^ 


*"**- 


o 








i— 1 


S3 


CM 


CM 


o 




o 


o 


o 












-^ 


CN 


00 


vO 


o 


o 


o 












o 


*-*- 


•~^» 


<-i 


«* 


r-. 


n 


1 










o 


O 


o 


■*^ 






o 




| 








o 


m 


p- 


CM 


03 


CO 


oo 


o 


^^ 








r-~ 


CO 


I— 1 


CO 


o 


1-3 


m 

o 
o 
o 

o 
o 


CN 


CN 




CM 


• 


o 












O 








CO 


rH 


o 












O 








CT* 


3 


o 




o 








O 






Q 


•— 1 


■"J 


co 




o 


O 


1 




•» 






^ 






^. 


o 


i-H 


o 


■^^ 




o 


1 




3 






o 


■^ 


**** 


o 


o 




o 


"^ 


1 






o 


o 


o 


CO 


o 




o 


O 


^» 


CO 






o 


o 


o 


"^» 


m 




n 


o 


o 


M 






r"» 


in 


m 


1 


i— i 




i— 1 


r— 1 


•— 1 


CO 
























M 
























> 
























53 


CO 


• 


u-i 




i 




o 










r-» 


4-1 


*o 


■—4 


^*» 


1 


<t 










>"5 


cr> 

i— 1 


O 

o 


CO 
CO 


1 


i 


1 


1 










ta 
























O 
























CO 
























s 


r-» 


• 




















O 


r^- 


4-1 


o 


CM 


i 


1 


1—1 










M 


o\ 


O 


in 


^-^ 


"*^» 


**^» 


^-^ 










H 


i— i 


o 


o> 


I 


i 


1 


1 










3 
























& 
























Pli 
























O 


vD 


• 


o 


O 






m 










PM 


l-» 


> 


o 


i-H 


i 


1 


CO 












ff* 


o 


o 


■^ 


-**» 


**-*» 


**^ 










Q 


■— 1 


£3 


<t 


1 


i 


1 


I 










«i 
























H 
























m 
























Pn 
























o 


vO 






















co 


vO 






















w 


CTv 






















H 


i-H 
1 




o 
o 

o 


* 
















H 






co 


1 






I 










H 


co 




**-* 


^-^ 






*^^ 










CO 


vO 




o 


o 






o 










W 


CTN 

i-H 




o 
o 


o 
o 

i-H 


i 
I 


1 
1 


o 
o 

00 











§ 



m 
o 

S3 
M 

Q 

W 
W 

o3 
« 









•O 










M 










•i-i 










Xi 










o 






>. 




•H 






XI 




o- 


• 




o 




o 


a. 




o 




n 


co 


s*. 


pa 




H 




X 




>1 




T3 


o 


T> 


XI 


T3 


U 


o 


(V 


o 


01 


• r-l 


pq 


4J 


o 


i-H 


x 




O 


pq 


•H 


01 


•o 


o 




CO 


4-1 


<u 


IH 


cj 


4-1 


(0 


,* 


1 


g 


1 


60 


CO 


•o 


o 


T3 


• i-l 


to 


0> 


M 


01 


l-l 


s 


03 


pq 


03 


fa 





CO 


l-l 






o 


01 






03 


4-> 






M 


CO 






m 


u 






o 


CO 






S3 


01 


e 




M 


XI 


u 




P 


CO 


0) 




w 




H 




w 


T) 






03 


01 


T3 


c 


PQ 


l-l 


<u 


u 




• t-l 


.* 


0) 


w 


CO 


u 


H 


rt 


4-1 


CO 




pq 


1 


x> 


>-. 


M 


01 


I 


4-1 


CO 


oo 


>. 


o 


CO 


T3 


(0 


o 


o 


01 


U 


CO 


Ph 


13 


O 



CO 








P-4 








00 








G 








•H 








T3 








01 








0) 








u 








pq 


01 






-o 


l-l 






0) 


X 






T3 


CO 




T3 


c 


i-l 


T3 


^ 


3 


•H 


l-l 


■■-I 


o 


CO 


•i-l 


X 


03 


> 


X 


01 




CO 


0) 


4-1 


•» 




4-> 


CO 


co 


CO 


cd 


oo 


4-1 


4-1 


oo 


•r-l 


i-l 


CO 


• l-l 


U 


3 


•o 


u 


IH 


T3 




4-1 




< 


O 




M 




(3 


4-1 


01 


i— I 




CO 


CO 


CO 




a) 


CO 


4-1 


<r-» 


n 


a> 


o 


1 


o 


hJ 


H 


"•^ 






^-^ 


1 


II 


II 




>*- " 










03 


CO 


• • 


* 


O 


rJ 


>. 








01 








w 









26 



CO • 

00 > 

on o 

-t Z 



u 

ts o 

00 

on . 



3 





> 


ON (J 




3 


-i o 




>-> 




c* 


fe 




01 
3 
C 
• ■-l 

4-1 

c 
o 
o 


o 


r-~ • 




r~ u 


00 


On o 


z 


-t O 


o 

l-l 

H 
3 


v£> • 




!=> 


r^ > 




Ph 


on o 


1 


o 


-H Z 


UJ 


a, 




l-J 

pa 

2 


£ 

M 


4J 




Pn 


vO O 




o 


ON 




00 


o. 




W 


1 01 




H 


00 




0j 


vO • 




M 


ON 00 




H 


— 1 3 




CO 


•< 




W 





en 




Q 




OS 




H 


Ml 


CQ 


c 




■■H 


e> 


T3 


z 


01 


i— i 


0) 


Q 


M 


W 


pa 


w 


--^ 


as 


00 


CQ 


c 


1 


• H 


z 


>^ 


O 


i— < 


z 


pE| 



o 
o 



o 



o 













» 






I-i 












01 






01 












i-i 






> 












M 






o 












3 






I— 1 












O 


i-i 

01 




Pi 








>N 




•o 


f— i 


01 


c 


•o 






•o 




01 


4-1 


c 


01 


01 






"O 




J= 


4-1 


o 


-o 


4-1 i-l 






o 




oo 


CO 


4-1 


1— 1 


CO 0) 


>* 


S-. 


z 




• l-l 


H 


CO 


o 


o U 


T3 


T3 




c 


x: 




c 


o 


u u 


TJ 


•o 


>N 


t-i 


4-1 


00 


l-l 




-C 01 


O 


o 


(0 


01 


1 


C 


3 


c 


4-> CU 
1 1 


z 


z 


oo 


H 


01 

1— 1 


•H 
l-l 


H 


CO 
o 


oi E 


§ 


.* 


I 


01 


4-1 


01 


>N 


•H 


4-1 M 


u 


o> 


4-1 


10 


"O 


•a 


l-l 


•H O 


o 


CO 


3 


■H 


•l-l 


e 


•o 


01 


3 W 


l-l 


»— 1 


■—1 


£ 


l-l 


CO 


3 


1 


pa 


pa 


oa 


3 


pa 


3 


OS 



27 









H 








O 








PS 








CO 








fa 








,-i w 








fa H 








M < 








CO fa 








CO Q 








O 








fa > 








►J 








• fa 








3 fe 








OS 








fa 








H 




fa 




fa CO 




O 


o 


<J >H 






a 


<: 




H 


3 


PS 




M 


co 


H oo 




fa 


fa 


fa -H 


fa 


3 


CO 


3 


fa 


as 


M 


« 




o 


> 




fa 


s 


OS 




fa 




<J 




M 


fa 


•-> 




< 


o 




fa 


H 




a 


<! 




1/1 


o 


H 




w 




O 




1/3 


CO 


H 




r> 


H 






<J 


<! 






o 


o 





CO 

o 
o 

PC 
H 
fa 
S 



O -* 



6-S 


6~« 


6^S 


r~- 


O 


c-. o 


st 


CN 





O >* e~ CN 



<Jv ITl 1^. (M CM i-( N 



o 
fa 

Q OS 

fa 
o 

CJ 



O fa 

w w 
*4 H 



3 



n o 

fa fa 

H H 

& r> 

o o 

o o 

a a 



w 
fa 
fa 



$ 



H fa 

O fa 

IS O 
co 33 
I 
c3 O 
fa 
O 
fa 
fa 
fa 



OS 



a a 



2 



O P 

fa fa 
fa fa 
fa fa fa 

H-22 
fa H H 



B-9 

r— I 



00 



CN 








00 


fa 




• 


O^ 


a 




H 


•—i 


fa 




c_> 




!-) 




o 


fa 


r~- 




CO 


a 


CN 




CN 


E=> 




CN 




•-> 


a 


00 


a CN 




w 


CTi 


fa 00 


-* 


w 


f-H 


fa C^ 


i— i 


13 




ES -i 




H 


•* 


H 


a 


W 


>■" 


fa « 


o 


CQ 


fa 


« • 






fa 


> 


H 


H 


>-3 


H O 


o 


O 




o a 


PS 


PS 


O 


PS 


CO 


CO 


t— 1 


co m 


o 








fa 








fa 








fa 








O 03 








a h 









fa 
o 

H 



28 



CO 

H 



3 



W 

m 



CO 
CO 

w 
o 
u 

S3 
CO 

e> 

z 

M 
P-. 



o 
z 



ss 



o 

CO 

H 
M 

a 

o 

H 



00 



CO 



CO 

w 
ex 

H 

Pm 

< 

u 

H 

BE 

M 
Z 

PL. 

2 



S~8 



o 



O CN 

00 r-t 



8^8 

o 



o 



s~« 6^s *« 

o> O O 
<"1 -h csl 






O 



o 
o 

CM 



CO 

H 



s 





Q 












O 












o 












lb 










X 


H 




Z 






CO 


< 




S 






HH 


<J> 




w 




4-» 


6h 






H 


4J 


c 




Q 


Ch 




r^ 


3 


X 


W 


M 


>H 


3 


J3 


CO 


Z 


Z 


H 


.c 


4J 


w 


z 


H 


O 


01 


CO 


OS 


< 


< 


O 


M 


o 


Ph 


o 


U 


CO 


Cm 


Pm 



29 



co 

H 
<C 

►J 

<! 



W 

H 





OS 






o 


w 




►J 


u 




o 


3 

pa 




H 






<$ 






O 






U 






f* 






« 




> 




>-< 


M 




2 


to 


fa 


o 


-J 


» 




m 


o 




< 


orf 




H 


CO 

H 






pn 


|H 




o 


pa 
pa 




CO 


< 




H 


H 




PS 






o 






M 






w 






s 





fas 
u 

pa 



to 



fa 



OS 


60 


o 


J«S 


l-l 


v—' 


o 




U 


H 




S3 


H 


O 


<fl 


X M 


o 


w w 


CJ 


co rs 



CM 
II 


CM 
II 


.— i 

CN 

II 

S 


o 

u-i 


in 


o 
in 


ro 


CN 


CI 


in 

v£> 


in 


m 
cn 



.— < i— I CM 



m 



o 

csl 



in 



oo 



CN 



in 
oo 



CO 

o 



CN r-t 

— < o 



o 

CO 



o 
o 














CO 












u 






85 






►4 


CO 




Q O 


w 






w 




<2 H 


o 




)>; 


l-l 


OS 


Z 




W 


^4 


w 


e <! 


<! 


w 


fa 


Js 


w 


Z M 


I-H 


o 






22 


<! > 


OS 


a 


iJ 


i-J 


h^ 


H W 


<u 


sa 


►J 


l-l 


§ 


w O 


> 


<$ 


< 



30 



TABLE V 

SEX AND COLOR RATIOS 
OF JARVIS ISLAND CATS 



COLOR 



MALES 



FEMALES 



TOTAL 



NUMBER 
PERCENT 

BLACK 



2i 

25 



36_ 
33 



63 
58 



TABBY 



Li 

13 



17_ 
16 



29 



GRAY 



1 

0.7 



10 
9 



BLACK & 
WHITE 



2 
1.4 



1 
0.7 



1 
3 



COLOR? 



I 
1 



TOTAL 



51 
48 



56 
52 



108 
100 



MAJOR VEGETATION TYPES OF 
JARVIS ISLAND N.W.R. 



MAJOR VEGETATION TYPES 


PH 


TR1BULIS 


El 


SESUVIUM 


■ 


LEPTURIS 


ED 


BOERHAV1A 




FIGURE II VEGETATION OF JARVIS ISLAND 



\W Z' IfeO" I VI. 

JARVIS ISLAND 

Modified from H0.H98 








o'zz' r o- 



Sesuvium 
towers ' . ,' meadow 

A ,-' Bare 

* * .' r guano flats 



5ft ^ :>*. w 









Portulaca 
20 ft. 



o i i \ i 

i 1 1 i i 

One Nautical Mile. 



FIGURE III LANDMARKS OF JARVIS ISLAND (BRYAN 1942) 



JARVIS ISLAND 



N MAG N 




CORAL BERM 
TERN COLONIES 
QUARRY 



_L 



STATUTE MILE 



25 



20 



15 



10. 



FIGURE IV 

LOCATION OF COLLARED CATS DURING 
RADIO-TELEMETRY FIXES 




NO OF HOURS HUNTED PER DAY 



NO. OF CATS SHOT PER DAY 



/ 



y \ / 
\ / 




NO OF DAYS HUNTED 



FIGURE V 



AMOUNT OF TIME HUNTED 
AND THE NUMBER OF CATS SHOT 



ATOLL RESEARCH BULLETIN 
No- 283 



VEGETATION AND FLORA 
OF NUI ATOLL, TUVALU 



By 
c- d- woodroffe 



Issued By 
THE SMITHSONIAN INSTITUTION 
Washington, D- C, U.S.A. 
May 1985 



— I 

177°09'E 



7"12S 



Reef edge 



1 kilometre 



-6' 



Nanumea 



— I - 
178' E 



■ Niutao 



o Nanumanga 



"NUI 



-8'S 



» Vaitupu 
O Nukufetau 



- 



8'S- 



Funafuti 



Nukulaelae * 



-10" 



10' - 



Niulakita <= 



78' E 

J 



Meang 



Tokinivae 




7'12'S- 



<^- 



Motuhkiliki 

Motuliki <*<L Moro//*/ 2 
Piliaieve ) } 



Tenamoitoka j 



Unima 



Unimai 2\ 




Fenua Tapu 



7' 15$- 



177' 09 E 



Fig. 1. Nui Atoll showing reef islands 



VEGETATION AND FLORA 
OF Will ATOLL, TUVALU 

By 
c- d- woodroffe 

Introduction 

Tuvalu, formerly the Ellice Islands of the Gilbert and Ellice 
Islands until separation on 1 October 1975, is a particularly remote 
group of islands in the Central Pacific. There are nine islands, five 
of which are atolls and four reef- top islands on a reefal platform. The 
vegetation and flora of these islands have received little attention. 

Most of the botanical collections from Tuvalu have been concentrated 
from the main island, Funafuti. The plants and their uses were first 
described by Hedley (1896) . A collection of plants was made by Mrs 
Edgeworth David (1899) during her residence on the island in July and 
August 1897. as part of the Royal Society of London expedition to core the 
atoll. Plants were also collected by Halligan and Finckh in 1898 and 
their specimens and those of Mrs David were described by Maiden (1904) . 
Since that time, the plants of Niutao have been described by Koch (1961) 
who has deposited a collection at the Smithsonian Institution, and those 
of Nanumea have been described by Chambers (1975) who has deposited a 
collection at the BP Bishop Museum. 

Nui Atoll, to the north of the Tuvalu group, has received very little 
attention since it was sighted by Alvaro de Mendana in 1568. In this 
paper the vegetation and flora of the atoll are described. Mapping of 
the vegetation of the atoll was done stereoscopically from black and white 
vertical aerial photographs taken in 1971 at a scale of approximately 
1:10,000. Vegetation units, and where possible individual trees, were 
verified on the ground and the map updated during a visit to the atoll 
of two weeks in February 1982 and a collection of the plants was made. 
The specimens are deposited at the herbarium of the Department of 
Scientific and Industrial Research, Botany Division, Christchurch, New 
Zealand, and I am grateful to Dr W Sykes of the D.S.I.R. for identifi- 
cations, and to Professor F R Fosberg and Dr M-H Sachet of the Smithsonian 
Institution for further identifications. The work was undertaken as part 
of a Land Resources Survey of Tuvalu, funded by F.A.O./ U.N.D.P. contracted 
to the Department of Geography, University of Auckland. I wish to thank 
Dr Roger McLean, project co-ordinator, for his guidance and encouragement, 
Paul Holthus and Salwa Woodroffe for help in the field and the Island 
Executive Officer, Lafaele, and the people of Nui for their hospitality. 

North Australia Research Unit, Australian National University, 
P.O. Box 41321, Casuarina NT 5792, Australia. 



Nui Atoll 

Nui Atoll (lat.7°12 , S long.l77°10 'E) lies to the north of the Tuvalu 
group, being approximately 130 km south of Niutao and 167 km north of 
Nukufetau. It consists of a reef platform approximately 7 km from north 
to south and 3 km from west to east (Fig. 1) . There is a shallow central 
lagoon, and the reef islands are concentrated along the eastern rim of the 
atoll; the western rim being a bare reef flat, extending nearly 4 km in 
length. Twenty reef islands were identified and visited; these have a 
total surface area of 337 ha and vary from the largest, Fenua Tapu, the 
island on which the principal settlement is located, with an area of 
138 ha, to Unimai 2 with an area of less than 0.02 ha. 

The atoll is inhabited, the population in 1977 being estimated at 
about 650. The principal settlements are on Fenua Tapu, with a smaller 
temporary settlement on Meang. Other islands are visited by canoe, or by 
walking around the eastern reef flat at low tide. 

Nui experiences a warm, humid climate throughout the year. The 
prevailing winds are easterlies. It receives about 3000 mm of rainfall 
annually with more than 200 rainy days per year. 

The reef islands of Nui are generally sandy with varying amounts of 
humus incorporated into the sand. A cemented rubble conglomerate platform 
underlies many of the islands and is prominent on the eastern rim of the 
atoll and along much of the lagoonward shore. The only extensive coral 
shingle or rubble ridges occur to the northwest of Meang. 

Nui is interesting in that the people of Nui show a much closer 
relationship to the islands of Micronesia than do other Tuvaluans, despite 
there being three islands closer to Kiribati (Gilbert Is.). Tradition has 
it that early Samoan settlers on Nui were largely replaced by people from 
Tabiteuea and Beru in Kiribati. The language on Nui is much more closely 
related to the language of Kiribati than on other islands in Tuvalu and 
names for things, including plants, are often different from the names 
used in the rest of Tuvalu. For instance, the root crop Cyrtosperma is 
called 'babai 1 on Nui whereas it is 'pulaka' on the other islands of 
Tuvalu. Where it has been possible to establish local Nui names for plants 
these are reported below. 

Vegetation Units 

The flora of Nui consists of at least 86 species. The largest number 
of species occurs on Fenua Tapu; 83 species were observed, many of which 
were crops or ornamentals found around the village. The following distinct 
vegetation units could be recognised: 

Pemphis scrub 

Pemphis acidula forms dense thickets at several sites, particularly 
on the lagoonward shore, on Nui Atoll. Pemphis occurs either on the 
dissected conglomerate platforms of the reef islands, or on a substrate of 



medium angular coral rubble or coarse sand, where inundation is infrequent 
and there is a sporadic cover of desiccated algal nodules or an algal mat. 

Pemphis scrub is found on the dissected conglomerate platform along 
the lagoonward shore of Fenua Tapu where it is 4-5m tall and forms either 
a continuous fringe less than 30m wide along the coast, or a discontinuous 
fringe scarcely more than an individual shrub wide. Similar Pemphis scrub 
occurs on a rubble or sand substrate near Telaeleke on Fenua Tapu , and at 
the northeastern end of that island, on Tokinivae, Pongalei and Talalolae. 
Within the Pemphis scrub, individual shrubs are spaced 4-5m apart, becoming 
increasingly sparse in the island interiors. On Tokinivae Pemphis may 
reach 10m tall; on Pongalei 8-10m tall, and on Talalolae and Fenua Tapu 
it rarely exceeds 6m tall. 

Pemphis scrub on the lagoonward shores grows so densely that there 
are generally no other plants associated with it, except occasionally the 
fern Polypodium . However on Tokinivae, Pongalei and Talalolae sparser 
Pemphis scrub extends inland, over a substrate of fine coral rubble and 
pinkish sand with an algal mat veneer, or more usually a cover of small 
black algal nodules, and in these areas Pemphis is generally less than 2m 
tall and Tournefortia and Scaevola also occur with a ground vegetation of 
Fimbristylis , Lepturus and Cassytha . 

A scrub composed of Pemphis is a common element of the vegetation 
of islands in the Pacific. It forms a distinct coastal zone on the atolls 
of the Cook Islands (Linton, 1933; Stoddart, 1975) , the Tokelau Islands 
(Parham, 1971), and continues north through Kiribati (Luomala, 1953; 
Moul, 1957) , and the Marshalls and Marianas (Fosberg, 1960) . On Nui Atoll 
Pemphis scrub occurs on the lagoonward side of many of the reef islands 
where these are narrowest, and extends into the islands reaching towards 
the oceanward shore. Pemphis tends to be growing at a lower elevation 
than surrounding vegetation, where it is prone to occasional flooding, and 
the areas of Pemphis scrub probably represent old infilled inter-island 
channels. The area at Telaeleke on Fenua Tapu has Pemphis scrub reaching 
to within 30m of the oceanward shore separated from the sea by a sand 
ridge . 

Scaevola scrub 

A dense scrub of Scaevola sericea occurs as a fringe around most of 
the perimeter of the majority of the reef islands of Nui, and often extends 
inland, generally in association with Acalypha , beneath Coconut woodland. 
The fringe may develop on sandy or on coral rubble substrates. It is 
particularly well developed along the oceanward shore of the reef islands, 
where it characteristically forms a fringe 15-20cm wide, rarely exceeding 
4m tall, landward of which Coconut woodland is found with the outermost 
coconuts overhanging the Scaevola . 

At the northeast of Fenua Tapu Scaevola scrub forms a belt approxi- 
mately 20m wide on a beach ridge of fine coral rubble and sand. This is 
a recently formed ridge which links what was previously the island of 
Tutupe to Telua, Fenua Tapu. 



Over much of the atoll Scaevola scrub is monospecific and the dense 
fleshy branches of Scaevola make the scrub penetrable only with difficulty. 
The creepers Canavalia and Cassytha where they occur on the scrub also 
impede passage. In some places Tournefortia , Cordia , Pandanus or 
Guettarda may be emergent while Triumfetta occurs on the ground. 

In addition to the coastal Scaevola scrub, Scaevola is also an 
important element of the inland scrub vegetation. A sparse Scaevola 
scrub, consisting of shrubs of Scaevola rarely exceeding 3m tall and 
emergent Pandanus , occurs on a series of north-south ridges of fine 
angular coral rubble to the northwest of Meang. A patchy ground cover of 
Boerhavia , Fimbristylis , Nephrolepis and Polypodium is found, and much of 
the vegetation is shrouded in Cassytha . Towards the interior of Meang, 
Tournefortia , Acalypha , Guettarda and Pisonia become more important and 
the sparse Scaevola scrub gives way to Scaevola/Acalypha scrub. 

Much of the interior of reef islands on Nui has a scrub vegetation 
composed principally of Scaevola sericea and Acalypha amentacea var. On 
Fenua Tapu Scaevola is the main element of the scrub, 3-4m tall, with 
Pipturus , Ficus , Guettarda , Morinda , Nephrolepis , Polypodium and 
Fimbristylis present and Acalypha only locally important. Elsewhere, as on 
Pongalei, Acalypha is the most conspicuous element of the scrub and 
Tournefortia , Asplenium and Boerhavia are also common. 

Pandanus is an important component throughout this vegetation unit. 
It is infrequent in most of the scrubland of Fenua Tapu, though becoming 
more common in the Scaevola/Acalypha scrub to the eastern end of the 
island. On Pongalei Pandanus is abundant within the Scaevola/Acalypha 
scrub and on Meang it is extremely common. These scrub areas are import- 
ant for the collection of Pandanus leaves, and the general increase in 
occurrence of Pandanus with distance from the village may reflect decreas- 
ing collection intensity. 

The Scaevola/Acalypha scrub is also found beneath Coconut woodland 
and beneath Pisonia woodland. Scaevola , and to a lesser extent Acalypha , 
grows best where there is plenty of light and is not well developed under 
dense woodland. Scrub is particularly important in areas of young Pisonia 
under trees up to 16m tall. 

Scaevola scrub forms a seaward belt on other islands in Tuvalu and 
has been described from Funafuti (Hedley, 1896) and over much of Nanumea 
(Chambers, 1975). It forms a prominent beach crest facies in the Tokelau 
Islands (Parham, 1971) on Swains Island (Whistler, 1983) , and on Onotoa, 
Kiribati (Moul, 1957) and is one of the most consistent vegetation units 
throughout Pacific atolls (Fosberg, 1953) . 

Tournefortia scrub 



Tournefortia argentea is generally taller than, and forms a more 
penetrable scrub than Scaevola . Small stands of Tournefortia occur within 
Scaevola scrub, and larger stands form a distinct scrub unit at the northen 
and southern ends of Pongalei, at the western end and to the southwest of 
Tokinivae and at the head of inlets on the ocean side of Pakantou and 
Unimai. 



The most extensive fringe of Tournefortia scrub however occurs 
along the sandy beach crest of the western shore of Meang. Here the 
scrub is 15-25m wide, and reaches 6-8m tall. Scaevola is found within 
the unit, and stands of Scaevola and Tournefortia alternate along the 
coast of central Meang, with replacement of Tournefortia by Scaevola 
scrub on the coarser substrate to the north of Meang. The Tournefortia 
scrub is much more open than Scaevola scrub and individual trees are 
spaced 6-8m apart. Canavalia , Triumfetta and Boerhavia occur within the 
scrub, while coconut, Guettarda and P andanus are occasionally emergent. 

Small pockets of Tournefortia scrub also occur inland. These are rarely 
extensive and consist of only one or a few individuals up to 17m tall. 
The vegetation within these pockets is usually typical Scaevola/Acalypha 
scrub or Pipturus/Acalypha/Scaevola scrub, sometimes with Pisonia . 
Similar Tournefortia scrub is an important littoral vegetation of many 
atolls in the Pacific, including the Tokelau Islands, Marshall Islands, 
Caroline Islands and northern Cook Islands (Linton, 1933; Fosberg, 1960; 
Niering, 1961; Parham, 1971) . 

Pipturus/Acalypha/Scaevola scrub 

Pipturus/Acalypha/Scaevola scrub is an inland scrub similar to the 
Scaevola/Acalypha scrub, but may be distinguished by the dominance of 
Pipturus argenteus in the upper storey. Pipturus grows to 10m or more 
tall, and often has several trees of Tournefortia in association, and 
occasionally Pisonia also. The lighter colour of the canopy of these 
emergent species allows recognition of this unit on aerial photographs. 
The lower storey vegetation is usually species-rich, with Scaevola , 
Acalypha , Ficus , Guettarda , Pandanus and the ferns Polypodium , Asplenium , 
Nephrolepis and Pteris . Pipturus/Acalypha/Scaevola scrub is found on 
Motupuakaka, Talalolae and Tokinivae, and to a lesser extent on Fenua 
Tapu. 

Rhizophora scrub 

Rhizophora stylosa is not extensive on Nui and is found only on the 
lagoonward shore of Fenua Tapu at Telaeleke. Here it rarely exceeds 4m 
in height, and forms a fringe around the edge of the conglomerate platform. 
This fringe is often only one tree wide, exceptionally reaching a belt 40m 
wide. Rhizophora scrub is monospecific and is backed by Pemphis scrub. 
Rhizophora and Pemphis may be interspersed where the two units are juxta- 
posed. Rhizophora scrub on Nui is more open than that described from a 
basin on Vaitupu (Woodroffe and Moss, 1984) . Cracks and fissures in the 
surface of the reef flat in this mangrove area were observed to flood and 
drain the area with the tides. 

Lumnitzera scrub 

The red-flowered mangrove Lumnitzera littorea occurs in two isolated 
pockets, each less than 20m x 30m wide, in the Pemphis scrub area of 
Telaeleke, Fenua Tapu. Lumnitzera reaches 3. 5-4. 5m tall, and each pocket 
is surrounded by Pemphis which is the only species recorded in association 
with Lumnitzera. 



Morinda thicket 

A thicket, dominated by Morinda citrifolia , occurs at one location 
surrounding a muddy depression to the east of Pongalei. The stand is 
only about 10m x 25m wide and is composed of Morinda 8m tall, spaced 
approximately 5m between individuals. The only associated plants are 
Ficus and the fern Asplenium . 

Pandanus grove 

Small groves of Pandanus tectorius occur around the coast of the 
reef islands of Nui, but do not form Pandanus woodland like that 
recorded on atolls, such as Aitutaki and Palmerston, in the Cook Islands 
(Stoddart, 1975; Sykes, 1976) , or Kayangel in the Palau Islands 
(Gressitt, 1952) . The groves are generally composed of 5-10 individuals 
which are 8-10m tall, though Pandanus on Fenua Tapu can reach as much as 
18m tall. 

Coconut woodland 

The most important woodland type on Nui is dominated by the coconut 
Cocos nucifera . The coconut palms exhibit a great variation in height 
and density, reaching 26m in some places. In those areas, as around the 
principal settlement, where there is regular collection of drinking nuts, 
the woodland is kept relatively clear of undergrowth while elsewhere it 
may be unattended and entirely overgrown with scrub. Such scrub tends 
to be dominated by either Scaevola or Acalypha , with Pandanus , Nephrolepis , 
Morinda , Guettarda , Ficus and Polypodium . Elsewhere Pisonia and Asplenium 
occur within the coconut woodland, or Asplenium alone may form a dense 
carpet between the palms. 

Pisonia woodland 

Woodland of Pisonia grandis is the most extensive natural woodland 
on Nui and occurs over much of the interior of the various reef islands, 
occasionally being exposed on the coast. It is well developed even on 
some of the smaller islands such as Teitai and Tenamoitoka. 

The most impressive stands of Pisonia woodland occur on Unimai, 
southern Meang and on Fenua Tapu at Te Kolokolo and near Telaeleke, in 
the latter two instances in association with deposits of phosphate. The 
massive Pisonia of Unimai reach heights of 22-24m and the woodland is 
dominated by immense individuals which exceed lm in diameter of the trunk. 
Acalypha is prominent in the understorey, with Ficus and rare Guettarda 
and Pandanus . The fern Asplenium , of which the young fronds are eaten as 
a spinach, is also an important associate in Pisonia woodland occurring 
both epiphytically on the Pisonia trees and on the coral rubble and sand 
substrate. On Meang similar massive Pisonia trees occur more sparsely, 
giving the woodland a more open appearance. In addition other large trees, 
including Hernandia , Calophyllum and occasional breadfruit Artocarpus are 
found in the woodland; the understorey vegetation is dominated by 
Asplenium but also contains Acalypha , Ficus and Polypodium. 



The Pisonia woodland at Te Kolokolo and Telaeleke on Fenua Tapu is 
also composed of large individual trees, many up to 20m tall; again 
Ficus , Morinda and Acalypha are important. The ferns Nephrolepis and 
Polypodium are found on the ground, however Asplenium is less abundant 
perhaps because the site is closer to the village. The Te Kolokolo area 
in particular has been altered as a result of felling of Pisonia trees, 
though the ability of Pisonia to shoot up from fallen limbs means that 
there has also been some regrowth. 

Elsewhere stands of Pisonia woodland are composed of less massive 
individuals often exceeding 18m in height but rarely more than 50cm 
diameter, and coconuts are more frequent. Such a Pisonia woodland is a 
typical woodland of many atolls in the Pacific including Kiribati, the 
Tokelau Islands, Cook Islands and Caroline Islands (Moul, 1957; 
Parham, 1971; Stoddart, 1975; Marshall, 1975), and is frequently 
associated with phosphatic substrates (Fosberg, 1953) . 

Hernandia woodland 

Woodland dominated by Hernandia sonora occurs at several sites just 
east of the pig wall on Fenua Tapu. Here large individual trees of 
Hernandia reaching 18m tall dominate small stands of woodland. Occasion- 
ally Pisonia may be present, and towards the north coast of Fenua Tapu 
Cordia is also associated with the Hernandia woodland. In view of the 
use of timber of Hernandia and its speed of growth, these groves of 
Hernandia located close to the village, and the individual trees in and 
around the village, have probably been planted. In a stand of Hernandia 
woodland on the central road of Fenua Tapu Acalypha is abundant, and 
Ficus , Guettarda and Asplenium are all frequent. Present but less common 
are Pipturus , Nephrolepis and Polypodium . 

Hernandia woodland also occurs extensively on Meang, though over much 
of southeastern Meang Hernandia and Pisonia grow in association forming 
a mixed open woodland more than 20m tall. Hernandia and Pisonia are 
closely associated in forests on Nanumea (Chambers, 1975) and have also 
been described together on Swains Island (Whistler, 1983) and on Aitutaki 
and Palmerston in the Cook Islands (Stoddart, 1975; Sykes, 1976) . 

Pulaka pits 

Small pits of pulaka Cyrtosperma chamissonis (locally called Babai) 
occur on a number of islands (Small, 1972) . One is found on Piliaieve, 
and several, less than 10m in diameter, occur on Tokinivae. The most pits, 
and those most important for production of pulaka, presently occur around 
the village on northwestern Fenua Tapu. Here pit construction is still 
underway and pits are regularly cultivated. In addition to pulaka, there 
is talo Colocasia , and banana Musa ; there are also several common weeds, 
most notably Ludwigia , Cyperus and Alternanthera . 

A further pulaka pit occurs on Meang. This is presently largely 
abandoned and pulaka grows only around the edge and in smaller secondary 
pits. Most of the pulaka pit is covered with Paspalum distichum with 
patches of Cyperus . Scaevola is found around the pit and grows with Ficus , 
Guettarda and Polypodium on small islands in the pit. 



Village and Gardens 

Natural vegetation has been almost totally replaced in the areas in 
which population is concentrated. The village is dominated by useful 
trees, particularly by the breadfruit Artocarpus , drinking coconuts, as 
well as scattered Hernandia trees. Additional areas, such as the hospital, 
guest house, and cemeteries have largely been planted with ornamental 
species, and these areas also support several weedy species. Ornamental 
garden plants include Pseuderanthemum atropurpureum , Clerodendrum inerme , 
Polyscias guilfoylei , Lantana camara , Plumeria rubra , Gardenia taitensis , 
Acalypha wilkesiana and Mirabilis jalapa . 



The flora 

The vascular plants collected or sighted on Nui Atoll are listed 
below. Numbers refer to voucher specimens deposited at the DSIR, 
Christchurch. 

ASPLENIACEAE 

Asplenium nidus L. [local name - laukatapa] 

Fenua Tapu: Woodroffe 189 ; Motupuakaka: sight; Pongalei: sight; 
Kokole: sight; Unimai: sight; Tenamoitoka: sight; Piliaieve: sight; 
Motuliki: sight; Motuliki 2: sight; Motulikiliki: sight; Talalolae: 
sight; Teitai: sight; Tokinivae: sight; Meang: sight; Nusafe: sight. 

DAVALLIACEAE 

Nephrolepis acutifolia (Desv.) Christ. [local name - lautamatama] 

Fenua Tapu: Woodroffe 140, 188 ; Motupuakaka: sight; Pongalei: 
sight; Kokole: sight; Unimai: sight; Tenamoitoka: sight; Piliaieve: 
sight; Pakantou: sight; Sikaiana: sight; Motuliki: sight; 
Motulikiliki: sight; Talalolae: sight; Teitai: sight; Tokinivae: sight; 
Meang: sight; Tokinivae 2: sight. 

Nephrolepis saligna Cass. [local name - lautamatama] 

Fenua Tapu: Woodroffe 187 

POLYPODIACEAE 

Polypodium scolopendria Burm.f . [local name - maile] 

Fenua Tapu: Woodroffe 120 ; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Unimai 2: sight; Tenamoitoka: sight; Piliaieve: sight; 
Pakantou: sight; Sikaiana: sight; Motuliki: sight; Motuliki 2: sight; 
Motulikiliki: sight; Talalolae: sight; Teitai: sight; Tokinivae: sight; 
Meang: sight; Tokinivae 2: sight. 

PSILOTACEAE 
Psilotum nudum (L.) Beauv. 

Fenua Tapu: Woodroffe 112 



PTERIDACEAE 
Pteris tripartita Sw. Ilocal name - te laukimoa] 

Fenua Tapu: Woodroffe 119 ; Unimai: sight; Meang: sight. 

PANDANACEAE 

Pandanus tectorius Park, (s.l.) [local names - teou, teto] 

Fenua Tapu: sight; Motupuakaka : sight; Pongalei: sight; Kokole: 
sight; Unimai: sight; Unimai 2: sight; Tenamoitoka: sight; 
Piliaieve: sight; Pakantou: sight; Sikaiana: sight; Motuliki: sight; 
Motuliki 2: sight; Motulikiliki: sight; Talalolae: sight; Teitai: sight; 
Tokinivae: sight; Meang: sight; Nusafe: sight; Tokinivae 2: sight; 
Tokinivae 3: sight. 

CYCADACEAE 
Cycas circinalis L. 

Fenua Tapu: Woodroffe 185 

GRAMINEAE 
Bambusa sp. 

Fenua Tapu: Woodroffe 168 

Cenchrus echinatus L. 

Fenua Tapu: Woodroffe 113 , 196 

Digitaria pacifica Stapf 

Fenua Tapu: Woodroffe 127 

Eleusine indica (L.) Gaertn. 

Fenua Tapu: Woodroffe 103, 124 

Eragrostis tenella (L.) P.Beauv.ex Roem. & Schult. 
Fenua Tapu: Woodroffe 110 ; Piliaieve: sight. 

Lepturus repens (Forst.f.) R.Br. 

Fenua Tapu: sight; Motupuakaka: sight; Pongalei: sight; Kokole: 
sight; Unimai: sight; Unimai 2: sight; Tenamoitoka: sight; 
Piliaieve: sight; Pakantou: sight; Sikaiana: Woodroffe 131 ; 
Motuliki: sight; Motulikiliki: sight; Talalolae: Woodroffe 180, 181, 182 ; 
Tokinivae: sight; Meang: sight; Nusafe: sight; Tokinivae 2: sight; 
Tokinivae 3: sight. 

Paspalum distichum L. 

Fenua Tapu: Woodroffe 147 ; Meang: sight 

Saccharum officinarum L. 
Fenua Tapu: sight. 



10 

Stenotaphrum micranthum (Desv.) C.E.Hubb. 

Fenua Tapu: sight; Motupuakaka: sight; Pongalei: sight; 
Tenamoitoka: sight; Pakantou: sight; Motulikiliki: sight; 
Talalolae: Woodroffe 179 ; Tokinivae: sight; Meang: sight. 

Thuarea involuta (Forst.f.) R.Br. 
Fenua Tapu: Woodroffe 109 

CYPERACEAE 
Cyperus alternifolius L. 

Fenua Tapu: sight; Meang: sight 

Fimbristylis cymosa R.Br. 

Fenua Tapu: Woodroffe 111,194 ; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Piliaieve: sight; Pakantou: sight; Tokinivae: sight; 
Meang: sight; Tokinivae 2: sight 

PALMAE 

Cocos nucifera L. [local name - Niu] 

Fenua Tapu: sight; Motupuakaka: sight; Pongalei: sight; Kokole: 
sight; Unimai: sight; Unimai 2: sight; Tenamoitoka: sight; Piliaieve: 
sight; Pakantou: sight; Sikaiana: sight; Motuliki: sight; Motuliki 2: 
sight; Motulikiliki: sight; Talalolae: sight; Teitai: sight; 
Tokinivae: sight; Meang: sight; Nusafe: sight; Tokinivae 2: sight; 
Tokinivae 3: sight. 

ARACEAE 
Colocasia esculenta (L.) Schott [local name - talo] 

Fenua Tapu: sight; Tokinivae: sight; Meang: sight 

Cyrtosperma chamissonis (Schott) Merr. [local name - babai] 

Fenua Tapu: sight; Piliaieve: sight; Tokinivae: sight; Meang: 
sight. 

AMARYLLIDACEAE 
Crinum asiaticum L. [local name - te luhe] 

Fenua Tapu: Woodroffe 135 

TACCACEAE 
Tacca leontopetaloides (L.) O.Ktze. [local name - masua] 

Fenua Tapu: Woodroffe 115 

MUSACEAE 
Musa sp. [local name - ulu] 

Fenua Tapu: sight; Meang: sight. 



11 

CASUARINACEAE 
Casuarina equisetifolia L. 

Fenua Tapu: Woodroffe 171 

MORACEAE 

Artocarpus altilis (Park.) Fosb. [local name - mei] 

Fenua Tapu: Woodroffe 193 ; Pongalei: sight; Tokinivae: sight; 
Meang: sight. 

Ficus prolixa Forst. f . 

Fenua Tapu: Woodroffe 122 

Ficus tinctoria Forst. f. {local name - pelo] 

Fenua Tapu: Woodroffe 139 ; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Tenamoitoka: sight; Piliaieve: sight; Pakantou: sight; 
Motuliki: sight; Motuliki 2: sight; Motulikiliki: sight; Talalolae: 
sight; Teitai: sight; Tokinivae: sight; Meang: sight; Nusafe: 
sight; Tokinivae 2: sight; Tokinivae 3: sight. 

URTICACEAE 

Laportea interrupta (L.) Chew 

Fenua Tapu: Woodroffe 195 ; Pongalei: Woodroffe 133; Tokinivae: 
sight; Meang: sight 

Pilea microphylla (L.) Lieb. 

Fenua Tapu: Woodroffe 186 

Pipturus argenteus (Forst. f.) Wedd. {local name - te pau] 

Fenua Tapu: Woodroffe 136 ; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Tenamoitoka: sight; Piliaieve: sight; Pakantou: 
sight; Motuliki: sight; Motulikiliki: sight; Talalolae: sight; 
Teitai: sight; Tokinivae: sight; Meang: sight; Nusafe: sight; 
Tokinivae 2: sight; Tokinivae 3: sight. 

OLACACEAE 

Ximenia americana L. [local name - kanana] 

Fenua Tapu: Woodroffe 129; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Tenamoitoka: sight; Piliaieve: sight; Pakantou: sight; 
Talalolae: sight; Tokinivae: Woodroffe 161 ; Meang: Woodroffe 157 

AMARANTHACEAE 
Achyranthes aspera L. approaching velutina H.SA.f local name - sisi vao] 

Fenua Tapu: Woodroffe 126 ; Motupuakaka: sight; Tokinivae: sight; 
Meang: sight. 

Alternanthera sessilis (L.) R.Br. 

Fenua Tapu: Woodroffe 150 



12 



NYCTAGINACEAE 

Boerhavia tetrandra Forst. 

Fenua Tapu: Woodroffe 104, 130 ; Motupuakaka: sight; Pongalei: sight; 
Piliaieve: sight; Pakantou: sight; Sikaiana: sight; Motuliki: sight; 
Motulikiliki: sight; Talalolae: sight; Tokinivae: sight; Meang: sight; 
Tokinivae 2: sight; Tokinivae 3: sight. 

Mirabilis jalapa L. [local name - petel] 

Fenua Tapu: Woodroffe 172 

Pisonia grandis R.Br. [local name - puka vai] 

Fenua Tapu: sight; Motupuakaka: sight; Pongalei: sight; Unimai: 
sight; Tenamoitoka: sight; Piliaieve: sight; Pakantou: sight; 
Motuliki: sight; Motuliki 2: sight; Motulikiliki: sight; Talalolae: 
sight; Teitai: sight; Tokinivae: Woodroffe 163 ; Meang: sight; Nusafe: 
sight. 

PORTULACACEAE 

Portulaca australis Endl. . 

Fenua Tapu: Woodroffe 197 

Portulaca lutea Sol. or P.oleracea L. 

Piliaieve: Woodroffe 101 ; Pakantou: sight; Tokinivae: sight; 
Meang: sight 

LAURACEAE 

Cassytha filiformis L. [local name - te louku] 

Fenua Tapu: sight; Motupuakaka: sight; Pongalei: sight; Unimai: 
sight; Piliaieve: sight; Pakantou: sight; Sikaiana: sight; Motuliki: 
sight; Motulikiliki: sight; Talalolae: Woodroffe 178 ; Tokinivae: sight; 
Meang: sight; Tokinivae 2: sight. 

HERNANDIACEAE 

Hernandia sonora L. [local name - puka] 

Fenua Tapu: Woodroffe 118 ; Pongalei: sight; Talalolae: sight; 
Tokinivae: sight; Meang: sight 

CRASSULACEAE 
Bryophyllum pinnatum (Lam.) Kurz (= Kalanchoe pinnata (Lam. ) Pers . ) 
Fenua Tapu: Woodroffe 138 

LEGUMINOSAE 

Canavalia cathartica Thou. [local name - lokou] 

Fenua Tapu: sight; Motupuakaka: sight; Pongalei: sight; Kokole: 
sight; Unimai: sight; Unimai 2: sight; Tenamoitoka: sight; Piliaieve: 
sight; Pakantou: sight; Motuliki: sight; Talalolae: sight; Tokinivae: 
Woodroffe 162; Meang: sight; Nusafe: sight; Tokinivae 2: sight. 



13 
Vigna marina (Burm.) Merr. {local name - te louku] 

Fenua Tapu: Woodroffe 107 ; Pongalei: sight; Unimai: sight. 

SURIANACEAE 
Suriana maritima L. [local name - ngie] 

Motupuakaka: Woodroffe 128; Pongalei: sight; Tokinivae: sight. 

EUPHORBIACEAE 

Acalypha amentacea Roxb. var. I local name - kakarapus] 

Fenua Tapu: Woodroffe 137; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Tenamoitoka: sight; Piliaieve: sight; Pakantou: sight; 
Motuliki: sight; Talalolae: sight; Tokinivae: sight; Meang: sight; 
Tokinivae 2: sight; Tokinivae 3: sight. 

Acalypha amentacea ssp. wilke.siana (Muell.-Arg.) Fosb. 
Fenua Tapu: Woodro ffe 169 

Euphorbia chamissonis (K1.& Gke) Boiss. 

Fenua Tapu: Woodroffe 134 ; Motupuakaka: sight; Pongalei: sight. 

Jatropha curcas L. 

Fenua Tapu: Woodroffe 184 

Phyllanthus amarus Schum. [local name - te uteute] 

Fenua Tapu: Woodroffe 105 

TILIACEAE 

Triumfetta procumbens Forst. f . [local name - kiaou] 

Fenua Tapu: Woodroffe 154 ; Motupuakaka: sight; Pongalei: sight; 
Kokole: sight; Tenamoitoka: sight; Piliaieve: sight; Pakantou: sight; 
Talalolae: sight; Meang: sight. 

MALVACEAE 
Sida fa 1 lax Walp. 

Fenua Tapu: Woodroffe 164 ; Tokinivae: sight 

GUTTIFERAE 

Calophyllum inophyllum L. [local name - itai] 

Fenua Tapu: Woodroffe 176 ; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Tenamoitoka: sight; Piliaieve: sight; Pakantou: sight; 
Talalolae: sight; Teitai: sight; Tokinivae: sight; Meang: sight. 

CARICACEAE 
Carica papaya L. [local name - esi] 

Fenua Tapu: Woodroffe 191 



14 

CUCURBITACEAE 
Cucurbita pepo L. 

Fenua Tapu: Woodroffe 19 

LYTHRACEAE 

Pemphis acidula Forst. n 

— I local name - ngie] 

.• v,/ e T, TaPU: Si9ht '' Pon 9 alei: si ght; Unimai: sight; Piliaieve- 
sight; Pakantou: sight; Motuliki: sight; Talalolae! sight 
Tokinivae: Woodroffe 160; Meang: sight; Tokinivae 2: sight! 

LECYTHIDACEAE 

Barringtonia asiatica L. n , 

[local name - ulu] 

TalalolIT TaP ^ W °° d " 0ffe 175; Pon galei: sight; Pakantou: sight; 
Talalolae: sxght; Tokinivae: sight; Meang: sight. 

RHIZOPHORACEAE 

Rhizophora stylosa Griff. r , 

"* I local name - te tongo] 

Fenua Tapu: Woodroffe 146 

COMBRETACEAE 

Lumnitzera littorea (Jack) Voigt n««»i 

y I local name - tangali] 

Fenua Tapu: Woodroffe 116 

Terminalia samoensis Rech n 

— ■ — I local name - te ipe] 

Fenua Tapu: sight; Motupuakaka: sight; Pongalei. sight- 

ONAGRACEAE 
Ludwigia octovalvis (Jacq.) Raven 
Fenua Tapu: Woodroffe 148 

CONVOLVULACEAE 
Ipomoea batatas (L. ) Lam. 

Fenua Tapu: Woodroffe 192 
Ipomoea roacrantha R. & s. 

Talalolae: Woodroffe 177 : Meang: sight 

BORAGINACEAE 

Cordia subcor data Lam. r , 

■ — I local name - kanava] 

Fenua Tapu: WoodroffeJL52; Pongalei: sight; Unimai- sight • 
sight lalSc^ Pak r° U: Slght; M ° tUliki = ***' ^ M: 

Sght; £E£r a rs* t Teitai: sight? Tokinivae: sight; — *■ 



15 

Tournefortia argentea L.f (= Messerschmidia argentea (L.f.) Johnst. 

= Argusia argentea (L.f.) Heine) 

[local name - tausunu] 

Fenua Tapu: Woodroffe 165 ; Motupuakaka: sight; Pongalei: sight; 
Unimai: sight; Piliaieve: sight; Pakantou: sight; Motuliki: sight; 
Motulikiliki: sight; Talalolae: sight; Teitai: sight; Tokinivae: sight; 
Meang: sight; Nusafe: sight; Tokinivae 2: sight; Tokinivae 3: sight. 

VERBENACEAE 
Clerodendrum inerme (L.) Gaertn, [local name - inato] 

Fenua Tapu: Woodroffe 167 

Lantana camara L. [local name - kai puaka] 

Fenua Tapu: Woodroffe 166 
Premna obtusifolia R.Br. [local name - te ango] 

Fenua Tapu: Woodroffe 153; Pongalei: sight. 

SOLANACEAE 
Physalis angulata L. [local name - te peen] 

Fenua Tapu: Woodroffe 145 

Solanum lycopersicum L. (= Lycopersicon esculentum Mill.) 
Fenua Tapu: Woodroffe 183 

ACANTHACEAE 
Pseuderanthemum carruthersii var. atropurpureum (Bull) Fosb. 
Fenua Tapu: Woodroffe 159 

RUBIACEAE 
Gardenia taitensis DC. [local name - siale] 

Fenua Tapu: Woodroffe 174 

Guettarda speciosa L. [local name - uli] 

Fenua Tapu: Woodroffe 121 ; Motupuakaka: sight; Pongalei: sight; 
Kokole: sight; Unimai: sight; Unimai 2: sight; Tenamoitoka: sight; 
Pakantou: sight; Sikaiana: sight; Motuliki: sight; Motuliki 2: sight; 
Motulikiliki: sight; Talalolae: sight; Teitai: sight; Tokinivae: sight; 
Meang: sight; Nusafe: sight; Tokinivae 2: sight; Tokinivae 3: sight. 

Hedyotis romanzof f iensis (Cham. & Schlecht Fosb. 

Pakantou: Woodroffe 132 ; Talalolae: sight; Tokinivae: sight; 
Meang: Woodroffe 156 (Western extension for species) 



16 

Morinda citrifolia L. [local name - te non] 

Fenua Tapu: Woodroffe 143 ; Motupuakaka: sight; Pongalei: sight; 
Kokole: sight; Unimai: sight; Tenamoitoka: sight; Piliaieve: sight; 
Pakantou: sight; Sikaiana: sight; Motuliki: sight; Motulikiliki: 
sight; Talalolae: sight; Teitai: sight; Tokinivae: sight; Meang: 
sight; Tokinivae 2: sight. 

APOCYNACEAE 

Neisosper ma oppositifolia (Lam.) Fosb. & Sachet [local name - pau pau] 

Fenua Tapu: Woodroffe 125 ; Pongalei: sight; Unimai: sight; 
Tenamoitoka: sight; Meang: sight. 

Plumeria rubra L. [local name - pua fiti] 

Fenua Tapu: Woodroffe 170 

ARALIACEAE 
Polyscias guilfoylei (Bull) Bailey 
Fenua Tapu: Woodroffe 142, 173 

GOODENIACEAE 

Scaevola sericea Vahl (=S . taccada (Gaer tn . ) Roxb . ) I local name - ngahu] 

Fenua Tapu: Woodroffe 158 ; Motupuakaka: sight; Pongalei: sight; 
Kok'ole: sight; Unimai: sight; Unimai 2: sight; Tenamoitoka: sight; 
Piliaieve: sight; Pakantou: sight; Sikaiana: sight; Motuliki: sight; 
Motuliki 2: sight; Motulikiliki: sight; Talalolae: sight; Teitai: 
sight; Tokinivae: sight; Meang: sight; Nusafe: sight; Tokinivae 2: 
sight; Tokinivae 3: sight. 

COMPOSITAE 
Adenostemma lanceolatum Miq. 

Fenua Tapu: Woodroffe 117 

Eclipta prostrata (L.) L. 

Fenua Tapu: Woodroffe 114 

Synedrella nodiflora (L.) Gaertn. 
Fenua Tapu: Woodroffe 123 

Vernonia cinerea (L. ) Less. 

Fenua Tapu: Woodroffe 106 

Wollastonia biflora (L.) DC. [local name - louku] 

Fenua Tapu: Woodroffe 108 



17 

Discussion 

The vegetation cover is more or less continuous over the reef islands 
of Nui Atoll, being composed principally of scrub or woodland of Pisonia 
or coconuts. There are no extensive areas bare of vegetation as on drier 
atolls, such as in Kiribati to the north. The sparse scrub communities 
occur largely on unfavourable substrates; the Scaevola scrub is sparse 
on the rubble ridges of north west Meang, and Pemphis scrub is sparse in 
low lying areas central to several reef islands which seem to represent 
old inter-island channels and which are probably liable to infrequent 
inundation. Suriana does not occur as a distinct scrub unit, as in the 
Cook Islands, but as isolated shrubs. 

The vegetation has been modified by human disturbance. Clearing of 
scrub from beneath coconut woodland is common practise around the village 
and was observed to have occurred on many other plots. Collection of 
domestically important plants almost certainly accounts for the present 
distribution of these plants; for instance the edible fern Asplenium 
is rare close to the settlement on Fenua Tapu, even in Pisonia woodland 
where it might be expected, and where it does occur is more commonly 
epiphytic than growing on the ground. It is much more common on islands 
further from the main village where it grows both epiphytically and on 
the ground. Pandanus leaves are also likely to have been collected 
intensively close to the settlements, and Pandanus is a much less obvious 
component of scrubland on Fenua Tapu than elsewhere. The main area for 
collection of Pandanus leaves was on Meang in 1982. 

The flora is composed of plants that have a widespread distribution 
throughout the Pacific and is similar to that recorded in Kiribati, 
the Tokelau Islands or on other islands in Tuvalu. Neither Thespesia 
populnea nor Hibiscus tiliaceus were recorded on Nui, though both occur 
on neighbouring Vaitupu and were collected on Nanumea to the north. 
Ximenia americana appears to be found only on the northern islands of 
Tuvalu, and has not been observed on the southern atolls. 

It is interesting that Acalypha amentacea var. is such a prominent 
element of the flora of the scrublands, growing with Scaevola . Acalypha 
is not recorded as such a common element of scrublands outside Tuvalu; 
it is frequent on Vaitupu, but is restricted to one occurrence on 
Nukulaelae where it was introduced for compost for the pulaka pits. 

There are a lot of introduced plants on Fenua Tapu, where garden 
crops, ornamentals and weeds have been introduced, but there are many 
less exotics than on Vaitupu or Funafuti. 

References 

Chambers, A. 1975 Nanumea Report. A socio-economic study of Nanumea 
Atoll, Tuvalu. Victoria University of Wellington. 

David, Mrs Edgeworth 1899 Funafuti or three months on a coral atoll: 
an unscientific account of a scientific expedition. London. 
John Murray. 

Fosberg, F.R. 1953 Vegetation of Central Pacific Atolls, a brief summary. 
Atoll Res. Bull. 23i 1-26 * 



18 

Fosberg, F.R. 1960 The vegetation of Micronesia. 1. General 

descriptions, the vegetation of the Marianas Islands, and a 
detailed consideration of the vegetation of Guam. 
Bull. Am. Mus. Nat. Hist . 119, 1-76.. 

Gressitt, J.L. 1952 Description of Kayangel Atoll, Palau Islands. 
Atoll Res. Bull . 14, 1-5. 

Hedley, C. 1896 General account of the atoll of Funafuti. 
Mem. Aust. Mus . 3, 1-71. 

Koch, G. 1961 Die materielle Kultur der Ellice-Inseln. Veroff . 
Museums fur des v8lkerkunde, Berlin. N.F.3, 199pp. 

Linton, A.M. 1933 Notes on the vegetation of Penrhyn and Manihiki 
Islands. J. Polynes. Soc . 42, 300-307. 

Luomala, K. 1953 Ethnobotany of the Gilbert Islands. Bishop Mus . 
Bull . 213, 129pp. 

Maiden, J.H. 1904 The botany of Funafuti, Ellice Group. P roc. Linn . 
Soc. N.S.W. 29, 539-556. 

Marshall, M. 1975 The natural history of Namoluk Atoll, Eastern 
Caroline Islands. Atoll Res. Bull . 189, 1-53. 

Moul, E.T. 1957 Preliminary report on the flora of Onotoa Atoll, 
Gilbert Islands. Atoll Res. Bull . 57, 1-48. 

Niering, W.A. 1961 Observations on Puluwat and Gaferut, Caroline Islands. 
Atoll Res. Bull . 76, 1-10. 

Parham, B.E.V. 1971 The vegetation of the Tokelau Islands with special 
reference to the plants of Nukunono Atoll. N.Z. Jl. Bot . 9, 
576-609. 

Small, C.A. 1972 Atoll agriculture in the Gilbert and Ellice Islands. 
Department of Agriculture , Tarawa, 154pp. 

Stoddart, D.R. 1975 Vegetation and floristics of the Aitutaki motus. 
Atoll Res. Bull . 190, 87-116. 

Sykes, W.R. 1976 Vegetation of Palmerston Atoll. Unpubl. Report 
Botany Division, DSIR, Christchurch. 

Whistler, W.A. 1983 The flora and vegetation of Swains Island. Atoll 
Res. Bull . 262, 1-25. 

Woodroffe, CD. and Moss T.J. 1984 Litter fall beneath Rhizophora stylosa 
Griff., Vaitupu, Tuvalu, South Pacific. Aquatic Botany , 18, 
249-255. 



Nusafe 



Meang 




[•>■:■:■:] Scaevola / Acalypha scrub 

||y[>:|:| Scaevola / Acalypha scrub with Pisonia 

R7J-R] Scaevola / Acalypha scrub with sparse 
:' << 'K -\ coconut 

Tournefortia scrub 
Coconut woodland 
Pisonia woodland 
Pisonia / Hernandia woodland 
Coconut woodland with Pisonia 
Pulaka Pit 
INDIVIDUAL TREES 

• Pisonia 
o Hernandia 
■ Calophyllum 

Barringtonia 

Cordia 

Tournefortia 



Fig. 2. Vegetation of Meang and Nusafe 



500 metres 







VEGETATION UNITS 

|||s>|s| Pemphis scrub 
fc\X\^ Sparse Pemphis scrub 
j Scaevola scrub 
J Scaevola / Acalypha scrub 
\[\- J Scaevola / Acalypha scrub with Pisonia 

'H Scaevola / Acalypha scrub with sparse coconut 
■Wig Tournefortia scrub 

p';VX^ Pipturus / Acalypha / Scaevola scrub , 

1 | Coconut woodland 

Pisonia woodland 
' . ' | Coconut woodland with Pisonia 
INDIVIDUAL TREES 

• Pisonia 
o Hernandia 
■ Calophyllum 
D Artocarpus 

* Tournefortia 
» Cordia 



Fig. 3. Vegetation of Tokinivae, Tokinivae 2 and Tokinivae 3 



Teitai 




Talalolae 



Motulikiliki 



Motuliki 2 



Motuliki 



Pakantou 




Piliaieve 



500 metres 



VEGETATION UNITS 

Pemphis scrub 
Sparse Pemphis scrub 
Scaevola scrub 
J Scaevola / Acalypha scrub 

Scaevola / Acalypha scrub with sparse coconut 
j Tournefortia scrub 
j^S'ij Pipturus / Acalypha / Scaevola scrub 
I | Coconut woodland 

f |'I"J Pisonia woodland 
| ' - ' -| Coconut woodland with Pisonia 
INDIVIDUAL TREES 

• Pisonia 

o Hernandia 

■ Calophyllum 

*■ Barringtonia 

* Cordia 

♦ Tournefortia 



Tenamoitoka 



Fig. 4. Vegetation of Teitai, Talalolae, Motulikiliki, Motuliki, 
Motuliki 2, Sikaiana, Pakantou, Piliaieve and Tenamoitoka 



Unimai 




Unimai 2 



500 metres 



VEGETATION UNITS 



Pongalei 




Pemphis scrub 
Sparse Pemphis scrub 
Scaevola scrub 
Scaevola / Acalypha scrub 
'•'•] Scaevola / Acalypha scrub with Pisonia 

j Scaevola / Acalypha scrub with sparse coconut 
BHHH Tournefortia scrub 
j I ; \ Morinda thicket 

Pipturus / Acalypha / Scaevola scrub 
J Coconut woodland 
1 • " • > Pisonia woodland 
I Coconut woodland with Pisonia 

't't'I Pandanus grove 
INDIVIDUAL TREES 

* Pisonia 

o Hernandia 

■ Calophyllum 

a Barringtonia 

* Cordia 

* Tournefortia 



Motupuakaka 



Fig. 5. Vegetation of Unimai, Unimai 2, Kokole, Pongalei 
and Motupuakaka 




3 
H 

C 
a) 



a) 
> 



bO 
•H 




a 

H 

C 

u 

<u 

4-1 

CD 

U 



XI 

3 

o 

CO 



£ 



4-1 

to 
1—1 

cu 




Plate 2. Scaevola scrub on recently formed beach ridge, northeastern 
Fenua Tapu 

Plate 3. Sparse Scaevola scrub, on fine rubble substrate, western 
Meang 





oo 




C 3 


3 P. 


O cd 


S^ H 


.G CO 


4-1 3 


•H C 


3 0) 


Pn 




-G C 


3 U 


U 0) 


O 4-1 


co to 


CO 


CO 


CD 


^H 







•* 


> 


XI 


0) 


C 


n 


3 


o 





u-j 


u 


■^ 


00 


CO 


01 


j: 


H 


CL 





>. 


U-l 


H 




CO 


c 


o 


•rl 


< 




~~ — 


CO 


en 


T3 


3 


)^ 


U 


cfl 


3 


4-J 


4-1 


(J 


&. 


0) 


•H 


3 


Ph 


O 


m 


0) 


4-1 


CO 





a. 




3 
Vj 

o 

CO 

CO 
•H 
4-1 
l-l 

o 

01 

3 
)-. 
3 
O 
H 



d 

0) 



o J= 

U CO 




B 

■H 
S 

3 



-a 
ti 
cfl 

iH 

■a 

o 
o 



■u 




B 




3 




•H 




c 




QJ 




H 




P. 




CD 




< 




& 




u 




■H 




& 






CD 


13 


CO 


C 


> 


n) 


■H 


^H 


C 


Td 


•H 


o 


^ 


o 


O 


13 


H 


4-1 


CI 


3 


M 


C 


OJ 


o 


4J 


o 


CO 


o 


a) 


o 


2s 


a) 




4J 




CO 




H 




cu 






r 







Plate 8. Scaevola/Acalypha scrub with coconuts, Unimai 



ATOLL RESEARCH BULLETIN 
No- 284 



INITIAL RECOLONIZATION OF FUNAFUTI 
ATOLL CORAL REEFS DEVASTATED BY 
HURRICANE "BEBE" 



By 
Hans Mergner 



Issued By 
THE SMITHSONIAN INST ITUIOW 
Washington, D- C, U.S.A. 
May 1985 







'Meteorol Service 
ir itrip 



▼ reef sections investigated 

routes to the reef sections 

<3=i direction of hurricane 'Bebe' 



INITIAL RECOLONIZATION OF FUNAFUTI 
ATOLL CORAL RELFS DEVASTATED BY 
HURRICANE "BEBE" 

By 
Hans Mergner 

Abstract 

On the 21st of October, 1972, hurricane "Bebe" devas- 
tated a large part of Funafuti atoll in the Ellice Islands. Among 
the most spectacular geomorphological alterations caused by the 
hurricane was a storm beach 19 km long, 4 m high and 37 m wide. 
The amount of coral debris washed up from the offshore coral reefs 
onto the reef flat was estimated at 2.8 x 10 tons of material 
(Baines, Beveridge and Maragos, 1974). The oceanside reef 
communities of the SE and E rim of the atoll had been totally 
destroyed, and those of the inner reefs of the lagoon side had been 
heavily damaged. Eight months after the storm a quantitative 
analysis of the resettlement and recruitment of coral species on 7 
reef sections was carried out: the destruction of the 
biophysiographic zones could be described as increasing from the 
northern border and also to the W rim of the atoll. Near the 
centre at Fongafale the lagoon reef flat was covered by thick 
carpets of the brown alga Dictyota bartaysii , possibly brought 
about by eutrophication effects. The resettlement of the reef flat 
by corals began with the recolonization of branching corals as well 
as regeneration of the very few surviving massive corals: about 80% 
of the number of new colonies belong to Acropora (mainly A. humilis 
and A^. hyacinthus ) , and about 20% to Pocillopora eydouxi , Porites 
lutea (?) and some Faviidae. The percentage of the area settled by 
the massive coral species is, however, greater than that settled by 
the branching species. Nevertheless, in the long-term, branching 
corals are expected to have a decisive influence on the future 
structural and biophysiographic zonation of the reef edge and reef 
flat, due to their more numerous young colonies, which are evenly 
scattered over the reef area, and due to their rapid growth rate. 
Consequently, an Acropora humilis - hyac in thus -community or an 
Acropora - Pocillopora eydouxi - assemblage can be predicted as the 
future biophysiographic zone. 



Institute for Special Zoology, Ruhr-University Bochum, 
D-4630 Bochum, F.R. Germany 



Introduction 

On the 21st of October, 1972, hurricane "Bebe" devastated a 
large part of Funafuti atoll, Ellice Islands (now Tuvalu) in the 
Southwestern Pacific. Six weeks later, December 10-24, 1972, 
Maragos, Baines and Beveridge (1973) investigated the 
geomorphological alterations caused by the cyclone especially on 
the SE side of the Atoll. A storm beach 19 km long, nearly 4 m 
high and 37 m wide was the most conspicuous geomorphological 
change. The amount of coral debris washed up from depths down to 
20 m and onto the reef flat was estimated at 2.8 x 10 tons of 
material and was derived from the outer reefs of this side which 
had been totally destroyed. In addition, the inner reefs of the 
lagoon side had also been heavily damaged. The authors gave a 
detailed report and some personal comments on the alterations which 
they found and the condition of the reefs at the Second 
International Coral Reef Symposium in Brisbane, 1973 (Baines, 
Beveridge and Maragos, 1974). These were used as a basis for the 
planned investigation on the recolonization of the destroyed reefs. 

Observations on the effects of tropical cyclones have been 
published by Blumenstock (1958, 1961) and Blumenstock et al. 
(1961), McKee (1959), Stoddart (1963, 1965, 1974), Tunnicliffe 
(1981), Woodley et al. (1981) and others. New growth and 
recolonization of corals have been described among others by 
Fishelson (1973), Loya (1976a, 1976b), Mergner (1979, 1981), 
Pearson (1981) and Schumacher (1977). Scientific descriptions of 
Funafuti atoll and its geology have been given by David and Sweet 
(1904), and as to the biology of the reef- forming organisms by 
Finckh (1904). 

In July (6-11), 1973, 8 1/2 months after the hurricane, and 
just after the Symposium, I carried out the first quantitative 
analysis of the recolonization of some reef sections of the atoll. 
For the purposes of comparison, several regions of the reef both 
flourishing and partly damaged, were investigated. The following 
is a report of these investigations. 



Seven sections examined on Funafuti reefs 

Funafuti atoll (Fig. 1) is located on the undersea ridge of 
the Ellice Islands about 1000 km north of the Fiji-Archipelago 
arising from a depth of more than 4000 m. It consists of 29 coral 
islands differing greatly in size, sometimes oblong in shape, 
sometimes round and very small, and covered with forest (Cocos 
nucifera, Pandanus tectorius , Plsonia grandis and others). The 
islands are connected by reef barriers and form a rectangle with 
one elongated corner (Fig. 2). Aside from numerous shoals, nine 
outlets (locally called "Te Ava") connect the open sea with the 
lagoon, which reaches a depth of 54 m at two positions and flattens 
to a few meters especially in the south. Some of the passages are 
navigable: in 1899 Agassiz (1903) had entered two of them on board 
the research vessel "Albatross". 

The Centre of hurricane "Bebe" struck the SE side of the atoll 
in October, 1972, and thus caused the greatest damage to the reefs 
located there (see Baines, Beveridge and Maragos, 1974). But the 
reefs of the NE side and of the S tip also suffered conspicuous 
damage in contrast to those of the SW and NW sides, which showed 
only few or nearly no effects of the storm because they had been 
better protected by the sheltering eastern reefs and the atoll 
lagoon. According to these criteria, the test areas and the reefs 
for comparison purposes had to be selected (Figs. 1, 12). 

Two sections were selected for the N rim of the destruction 
zone close to the N end of the isle of Tengako (sections 1 and 2), 
two for the S rim of this zone near the isle of Tutanga at the S 
end of the atoll (sections 6 and 7), two for the W side of the 
atoll lagoon near the isle of Faufatu at the W end of the atoll 
(sections 4 and 5) and one for the centre of the destruction zone 
near the E end of the main island Funafuti (section 3). Because of 
the high swell from the east and correspondingly high breakers, the 
exposed outer reef section could not be investigated. 



In all sections, the biophysiographic zonation was studied in 
a strip 20 to 100 m wide from the sea-shore across the reef flat to 
the fore reef or the lagoon floor. The distances from the mean 
water level to the reef edge were between 120 and 250m; greater 
distances could not be reached by snorkeling. 

In contrast to reef sections 4-7, which showed no or only 
little damage, nearly the whole reef flat of sections 1-3 had been 
destroyed. Here, in each case, a test area of 20 x 20 m or 20 x 10 
m was marked with plastic lines extending from the lagoon floor or 
reef slope over the reef edge onto the outer reef flat. Within the 
borders of each area all living remains of former coral colonies 
(without exception, e.g. massive or crus tose species) and newly 
settled colonies were marked on a map and their sizes were 
calculated by taking the average between the longest and shortest 
diameters. All drawings, measurements and underwater photographs 
were made by snorkeling down to a depth of 5 m. 

Results of zonation studies of reef sctions 1-7 

First, the structure, the flora and fauna and the 
biophysiographic zonation of all reef sections investigated will be 
briefly characterized. Then, the recolonization of the reef 
platforms will be analyzed using reef sections 2 and 3 as examples. 

Reef section 1 

Reef section 1 is situated at the N rim of the central 
destruction zone of hurricane "Bebe", and 300 m south of the N end 
of the isle of Tengako. It extends SW from the shore of the lagoon 
for 160-180 m with a minimum breadth of 20 m, and its zonation is 
shown in Table 1, Fig. 12. 

Reef section 1 shows serious damage to the reef platform and 
reef edge. Its biophysiographic zonation is characterized by 
sparse growth of different algae species on the abraded reef fiat. 
Within more than 500 m of the reef front, only 17 massive faviid 
colonies had survived, but no scleractinian colony had resettled. 
Aside from a few Pagurids and fishes no mobile fauna is visible. 

Reef section 2 

Section 2 (Fig. 3) runs parallel to and about 1200 m from 
section 1 and is located nearer the centre of the hurricane "Bebe" 
zone. It runs in a SW direction, is 140 m long and has a minimum 



breadth of 20 m. In many respects, its structure, settlement and 
zonation are similar to those of section 1. It does, however, 
differ from section 1 in that there is--aside from some remainders 
of dead coral colonies — a total lack of surviving living colonies. 
There are many species of fish in this area and, above all, there 
is a recolonization of numerous young scleractinian colonies (Table 
2, Fig. 12). 

Reef section 2 had been damaged to a greater extent by the 
hurricane than section 1: the old reef flat was largely eroded, the 
reef edge destroyed and its Acropora -zone totally demolished. No 
coral colony survived, but within 136 m of the reef front 84 young 
coral colonies had resettled. The fish fauna is much more 
plentiful than in section 1, both in numbers and in species. 

Reef section 3 

Reef section 3 is situated in the middle of the central 
destruction zone of hurricane "Bebe" , 120 m SW of the small jetty 
of Fongafale on Funafuti Island. It covers the inner reef of the 
lagoon side with a length of 120 m and a breadth of 20 m westward. 
Possibly due to the eutrophication by sewage of the surface water, 
which is only slightly agitated, a thick layer of the brown alga 
Dictyota bartaysii (Fig. 4) loosely covering the reef flat has 
developed. It largely conceals the serious damage due to the storm 
and allows only fragmentary insights into the settlement structures 
of this section. However, on alga-free coral rock areas along the 
reef edge large numbers of young stony coral colonies have 
resettled (Table 3, Fig. 12). 

Of all the reef sections investigated, section 3 was hit the 
hardest by the hurricane. Aside from damage caused by hurricane 
"Bebe", much of this area has been changed during World War II. In 
addition, large areas of the reef edge and the reef slope are 
covered with dense algal carpets that have probably resulted from 
eutrophication effects of sewage. Nowhere can uninjured surviving 
stony corals be found; living faviid colonies can only be found in 
limited areas. However, 67 coral colonies, mostly species of 
Acropora, have resettled on the alga-free areas along the reef 
edge, covering an area of nearly 50 m . The occurrence of numerous 
herbivorous and detritophagous fish species like Acanthuridae, 
Chaetodontidae, Mullidae and Pomacentridae probably should be 
attributed to the mass population explosion of the algal stocks. 

Reef section 4 

Reef section 4 (Figs. 5-7) extends eastward from the NE end of 
the horseshoe- like island of Faufatu in the middle of the W side of 
the atoll. It runs for a length of 120 m and a maximum breadth of 
100 m through the inner reef but does not reach the reef edge. In 



spite of its location facing the path of hurricane "Bebe", the 
storm damage—apart from some overthrown colonies of Acropora 
hyacinthus --is minimal, because the section is protected by an 
extensive reef platform located on its E side. This section is 
therefore suitable for comparison with the greatly damaged reef 
sections 1-3. Quantitative investigations of surviving and 
resettled young coral colonies were not necessary. Because of the 
breadth of section 4, it is differentiated into a northern (N) , a 
middle (M) , and a southern (S) strip in Table 4 when necessary 
(Table 4, Fig. 12). 

Generally speaking, reef section 4 is a suitable example of an 
undamaged inner reef with characteristic coral assemblages and 
biophysiographic zones. The reef edge with its coral communities 
at a distance of more than 800 m could not be reached by swimming 
because of the strong currents. However, a strong current of 25 
cm/s also occurred within the shore channel, and here similar 
assemblages, among others a well-developed Acropora hyacinthus 
(Pocillopora eydouxi )-zone, have settled. Their species 
composition should be comparable with those of the former zones of 
reef sections 2 and 3 before their destruction by the hurricane. 
Fishelson (1973), Loya (1976a) and Mergner (1981) have shown that 
damage caused by man-made perturbations and natural catastrophes 
will be gradually erased by regeneration and recolonization of the 
coral colonies if no further perturbation is added. Therefore, the 
expectation that the destroyed zones will gradually regain their 
former structure and species composition is also justified in this 
case . 

Reef section 5 

Reef section 5 begins at the curved south side of the isle of 
Faufatu, 250 m south of reef section 4. It covers the outer reef 
with a length of 120-180 m and a maximum breadth of 80 m to the SSW 
up to the surf zone (Table 5; Figs. 7, 12). 

This reef section shows the typical zonation of the southwest 
Pacific outer reef. Because of the strong surf, reef slope and 
fore-reef could not be inspected. The storm damage seems to be 
relatively small. Thus, reef section 5 is a good comparison for 
the destroyed outer reefs of the E side of the atoll. 

Reef section 6 

Reef section 6 (Fig. 9) also belongs to the outer reefs of the 
W side of the atoll. It begins near its southern end on the SW 
edge of the isle of Tutanga and runs westwards for 150 m with a 
breadth of 50 m to the upper reef slope. Because of its location 
on the lee-side of the path of hurricane "Bebe" and at the most 
extreme S edge of its destruction zone, the damage is minimal. The 
surface current was very rapid (40 cm/s) , but the surf of this reef 



site was weak during the investigation period. Therefore, a short 
study of the coral assemblages in the region between reef edge and 
fore reef was possible (Table 6, Fig. 12). 

Reef section 6 shows a characteristic boulder zone: the large 
coral blocks were apparently thrown up by storm waves of former 
cyclones coming from westerly directions. They are even cited in 
David and Sweet (1904). The coral assemblages of the greatly cleft 
reef edge and the steep reef slope with its deep spurs and grooves 
are very diversified. Both zones characterize this reef section, 
which combined with the zonation of the reef flat in section 5, 
gives an approximate impression of the original appearance of the 
outer reefs destroyed on the eastern atoll side. 

Reef section 7 

Reef section 7 (Figs. 9, 11) covers 160-200 m of the inner 
reef, is 50-70 m wide and runs from the SE corner of the isle of 
Tutanga towards the SE. Just as outer reef section 6, section 7 is 
influenced by a strong surface current of at least 25 cm/s flowing 
from the lagoon outwards to the open sea. In spite of its location 
at the S edge of the destruction zone of hurrricane "Bebe", no 
serious storm damage can be observed along this section. Possibly, 
the narrow south tip of the atoll lagoon did not provide a long 
enough fetch (only 1.5 km) with enough water (it is very shallow 
here) for the hurricane to build up large enough waves to cause 
extensive damage. Section 7 thus gives the impression, just as 
section 4, that it is an intact inner reef. Its biophysiographic 
zonation, however, differs in many respects from that of section 4 
and all other inner reef sections by the prevalence of microatolls, 
especially of Porites lutea , and by the lack of a characteristic 
reef slope (Table 7, Fig. 12). 

Seen as a whole, section 7 is of less interest for a 
comparison between undestroyed and destroyed inner reefs. 

Results of the quantitative analysis of reef sections 2 and 3 

All the inner and outer reef sections investigated on the W 
and SW side of Funafuti atoll (4-7) proved to be relatively 
undamaged and generally suitable models for comparison with the 
seriously damaged reefs of the E side. This is especially true for 
section 4 as an inner reef and a combination of sections 5 and 6 as 
model of an outer reef (Fig. 12). 

Thus, the reef flat of outer reef section 5, within the area 
of the abrasion zone, the shingle zone and the algal ridge, shows 
typical features of the zonation of SW Pacific atolls. In 
addition, its remaining coral zone in the region bordering on the 
algal ridge indicates the former existence of large stocks of the 
Halimeda - Montlpora f o 1 iosa - communi ty . However, the boulder zone on 
the reef flat, which is almost always present, is lacking in 



section 5, but well developed in outer reef section 6. The reef 
edge and upper reef slope of this section show the characteristic 
species composition of their coral assemblages. Therefore, a 
combination of the zonation of both sections probably offers a true 
representation of the sequence of zones on a W outer reef on 
Funafuti atoll. Unfortunately, this view could not be compared 
with the present aspect of the destroyed outer reefs on the E side. 
For details of their former aspect see the contributions of "The 
Atoll of Funafuti" (1904) and for the situation immediately after 
the hurricane see Baines, Beveridge and Maragos (1974). 

An impression of the later result of the recolonization of the 
eastern inner reefs can be formed much more easily by examining the 
coral zones of western inner reef section 4: these zones are 
characterized by Acropora humills- hyaclnthus and Pocillopora 
eydouxi -communitles , but not by massively growing faviids and 
poritids, which, in spite of early regeneration of the surviving 
colonies, are less important as predominant species compared with 
Acropora and Pocillopora. In contrast to that of section 4, the 
zoning of section 7 is less helpful for comparison with damaged 
inner reef sections 2 and 3, because it is characterized by 
microatoll formations especially of Pori tes lutea . Only isolated 
areas of the immediate reef edge are settled by typical 
Acropora - Pocillopora - communities. Conspicuous in both inner reef 
sections of the western side (4 and 7) is the almost complete lack 
of macroalgae on the reef flat, whereas on the inner reefs of the 
eastern side luxuriant algal stocks have developed. 

All three reef sections investigated on the E side of Funafuti 
atoll must be considered as largely damaged by storm, with the 
extent of damage increasing from the northern border (reef sections 
1 and 2) to the central destruction zone (section 3). An initial 
comparison of the three sections shows that, in the northernmost 
section a number of massively growing Plesiastrea colonies had 
indeed survived, but no new colonization was observed. On the 
other hand, in the more southerly sections 2 and 3, no colonies 
survived without damage and only smaller regenerating areas have 
arisen, but, at the same time, numerous young colonies of different 
size have initiated the first phase of recolonization (Figs. 13, 
14). Almost all of these pioneer species belong to Acropora, a few 
to Pocillopora and Porites, and some to the faviids. 

In order to verify results on species composition, species 
diversity, and settling density during the first phase of the 
recolonization, a careful quantitative analysis within the area of 
the reef edge and the close reef flat of both sections is 
necessary. 

Reef section 2 

Quantitative study of recolonization in reef section 2 (Fig. 
13) consisted solely of enumerations and measurements of the 



resettled and regenerated areas of the scleractinians within a 
strip 8 to 15 m wide behind the reef edge. In this area of about 
136 m , for reasons of simplification, all holes and channels with 
sandy bottom or debris were included, although these are usually 
not colonized. Thus, the area actually colonized should be reduced 
by about 25%. 

Within the area studied, 84 stony coral colonies were counted 
belonging to the following genera and species: 

Acropora (corymbosa , humllls , hyacinthus ) 67 colonies = 79.7% 

Goniastrea ( retiformlsl 12 colonies = 14.3% 

Platygyra ( lame 1 Una ) 1 colony = 1.2% 

Pocillopora ( damicornis , eydouxi ) 2 colonies = 2.4% 

Porites ( lutea ) - - ^ colonies = 2.4% 

Scleractinia 84 colonies =100.0% 

It is possible that some growth areas of Goniastrea originate 
from several regenerating segments of a formerly 
uniform larger colony, but this is not likely in the case of an 
Acropora humilis colony 30 cm in diameter. All remaining coral 
heads with smaller diameters may be considered as young colonies 
that settled during the 8 1/2 months after the passage of hurricane 
"Bebe" and since then have continued to develop, to reach varying 
and often considerable sizes. According to the time of settling 
the following diameters (0) and base areas of Acropora were 
determined: 



2 2 

6 colonies with up to 2 cm and 3.1 cm , altogether 18.6 cm„ 

14 colonies with 2.5- 3 cm and 7.1 cm , 

27 colonies with 3.5- 5 cm and 19.6 cm. 



10 colonies with 5.5- 8 cm and 50.2 cm. 

8 colonies with 8.5-10 cm and 78.5 cm 

1 colony with 10.5-15 cm and 176.6 cm 

1 colony with 15.5-30 cm and 706.5 cm 



alogether 99.4 cm 



altogether 529.2 cm 

altogether 502.0 cm 

altogether 628.0 cm 

altogether 176.6 cm„ 

altogether 705.5 cm 



2 
67 colonies with a settling area of altogether 2660.3 cm 

By contrast, only 12 young colonies or regenerating segments of the 

2 
massively growing Goniastrea inhabit 4979 cm , an area twice as 

large as that occupied by Acropora. The other 5 newly settled 

2 2 

colonies, covering 528 cm ( Platygyra with 314 cm , Pocillopora 

2 2 

with 157 cm and Porites with 57 cm ) are less significant for the 



recolonization of the reef area analyzed, which totals ca. 8168 

2 2 

cm . Of 136 m of the area concerned, the 84 young colonies oci 

only about 0.6%, which can be divided up as follows: 



10 

2 
Acropora with 67 colonies and 2660 cm area amounts to 0.20 % 

2 
Goniastrea with 12 colonies and 4979 cm area amounts to 0.37 % 

2 
Platygyra with 1 colony and 314 cm area amounts to 0.02 % 

2 
Pocillopora with 2 colonies and 157 cm area amounts to 0.01 % 

2 
Porites with 2 colonies and 57 cm area amounts to 0.004% 



2 
Scleractinia with 84 colonies and 8168 cm area amount to 0.60 % 

Goniastrea retlformis dominates all other coral species 
regarding its share of the settling area, but is not dominant in 
its significance for the future structural and biophysiographic 
zonation. The Acropora species, with their 67 young colonies 
equally distributed over this reef segment, are much more important 
and will influence the later appearance of the reef edge and the 
neighboring reef platform. There are two reasons for this: first, 
Acropora species grow much more rapidly than the massive colonies 
of Goniastrea, and are also able, because of their favorable 
distribution, to roof over the area of settlement within a 
relatively short period. Therefore, they are at an advantage over 
all the others in the competition for space and light. Second, 
they form a varied landscape with their diversified structures 
providing numerous hiding places and a good food source as the 
biocenosis for many different species of reef fauna. 

It can be assumed, and comparable observations from Red Sea 
coral reefs (Fishelson 1973, Loya 1976a, Mergner 1981) substantiate 
this assumption, that further undisturbed development will result 
in the establishment of an Acropora humllis- hyaclnthus -zone as the 
biophysiographic zone of the reef edge and neighboring reef flat, 
such as found in other reef regions of Funafuti atoll, as in reef 
section 4 for example (Fig. 7). 

Reef section 3 

In reef section 3 (Fig. 14), all newly-settled and 
regenerating stony corals were also enumerated, measured and 
exactly marked on underwater maps. However, here only a narrow 
strip, 2.5 m wide on the average along the reef edge, could be 
mapped, because large reef areas were covered by carpets of the 
brown alga Dictyota bartaysii (Fig. 4), and all new-settled stony 
coral colonies were found only on algal-free coral rock. In the 
test area of about 50 m , 67 young colonies were found belonging to 
the following genera (with species names): 

Acropora ( humilis , hyacinthus , pulchra etc.) 50 colonies = 74.6% 

Faviidae (Fayia, Goniastrea etc.) 7 colonies = 10.5% 

Pocillopora ( damicornis , eydouxi ) 7 colonies = 10.5% 

Porites ( lutea ) 3 colonies = 4.4% 

Scleractinia 67 colonies =100.0% 



11 



Only one favild colony can be assumed to be newly settled, the 
other 6 are regenerating parts of a former larger colony. Such an 
origin can also be assumed for one Porites lutea colony and 
possibly one for Acropora colony with a diameter of 20 cm. All the 
other coral heads are young colonies that have settled since the 
passing of hurricane "Bebe". The following diameters (0) and base 
areas corresponding to the time of settling have been ascertained 
for Acropora; 



3 colonies with up to 1 cm and 
9 colonies with 1.5- 2 cm and 



16 colonies with 2.5- 
14 colonies with 3.5- 

2 colonies with 5.5- 

4 colonies with 8.5- 

1 colony with 10.5- 15 cm and 176.6 cm 

1 colony with 15.5- 20 cm and 314.0 cm 



3 cm 

5 cm 

8 cm 

10 cm 



and 
and 
and 
and 



0.8 cm 

3.1 cm' 

7.1 cm 



19.6 cm 
50.2 cm 
78.5 cm 



altogether 

altogether 

altogether 
altogether 
altogether 
altogether 
altogether 
altogether 



2.4 cm 

27.9 cm" 

2 
113.6 cm, 

274.4 cm" 

100.4 cmj 

314.0 cm! 

176.6 cm, 

314.0 cm" 



50 colonies with a settling area of altogether 



1323.2 cm' 



By contrast, only 6 regenerating and one young faviid colony 
cover an area of 7870 cm , which is nearly 6 times more than 
Acropora ; Porites , which also grows massively, with only one 
regenerating and 2 young colonies, covers 648 cm , which is half 
that occupied by Acropora . The branched coral Pocillopora claims 
563 cm for 7 young colonies. Thus, the total area recolonized by 
corals amounts to 10403 cm . Based on 50 m of the reef area 
analyzed along the reef edge, the recolonization involves about 2%, 
which can be divided up as follows: 



Acropora with 50 colonies and 1323 cm area amounts to 0.26% 

2 
Faviidae with 8 colonies and 7870 cm area amounts to 1.57% 

2 
Pocillopora with 7 colonies and 563 cm area amounts to 0.11% 



Porites 



with 3 colonies and 648 cm area amounts to 0.13% 



In reef section 3 it is much more evident than in reef section 
2 that the area covered by the massive stony corals (Fayia, 
Goniastrea , Porites ) is much larger than that covered by the 
branched corals ( Acropora , Pocillopora ) . But regardless of this 
fact, the latter corals will definitely influence the future 
structural and biophysiographic zonation of the reef edge. The 
rapidly growing Acropora species will soon predominate in this reef 
area both numerically and physionomically and together with 
Pocillopora will build up a varied biocenosis for a reef fauna rich 
in species and numbers. A condition favoring this is the high 
initial settling density along the reef edge in section 3: on each 
square meter of free coral rock 1.14 young colonies of branched 



12 

corals settle, compared to only 0.5 colonies in reef section 2. It 
cannot be proved that this is due to the reduced amount of free 
settling area resulting from the unusual proliferation of algae. 
With further undisturbed development, an Acropora humilis- 
hyac in thus -zone or a mixed Acropora - Pocillopora -zone, will also 
develop in section 3 as the biophysiographic zone of the reef edge 
and neighboring reef platform. Estimates differ as to the time 
required for this development. However, further damage caused by 
new storms and/or by man's influence may change or delay or even 
prevent this evolution. 

Conclusions 

1. The geomorphological alterations of large parts of Funafuti 
atoll caused by hurricane "Bebe" have been described by Baines, 
Beveridge & Maragos (1974). Here some reef sections destroyed on 
the E side are compared with undestroyed sections on the W and SW 
side, and the first phase of the recolonization of two reef areas 
by stony corals is analyzed. 

2. The extent of the damage to the reef sections studied can be 
ascertained by the alterations of their structure and biophysio- 
graphic zonation: it increases from the northern border zone of the 
cyclone (reef sections 1 and 2) to the central destruction zone 
where section 3, close to Fongafale, had been most intensively 
struck. Wide areas of its devastated reef flat are covered by 
thick carpets of the brown alga Dlctyota bartaysll possibly because 
of eutrophication. 

3. Quantitative analyses in destroyed reef regions 2 and 3, show 
that recolonization by scleractinians is a result of regenerating 
segments of the few surviving massive coral colonies and of newly 
settling branched corals. Among them Acropora humilis and A. 
hyacinthus (and with reservations also A. corymbosa , formosa and 
pulchra ) and Pocillopora eydouxi are especially prominent: Acropora 
comprises 79.7% (sect. 2) to 74.6% (sect. 3) of the number of all 
young colonies, Pocillopora 2.4% (sect. 2) to 10.5% (sect. 3). 
However, the massive growth species outweigh the branched corals 
with respect to percentage of area settled; the massive growth 
species occupy 0.4% to 1.7% of the entire area, while the branched 
corals occupy only 0.2% to 0.4%. 

4. In spite of this, the branched corals will have a decisive 
influence on the future structural and biophysiographic zonation of 
the reef edge and the neighboring reef flat. This is indicated by 
the fact that their young colonies are much more numerous and 
evenly distributed over the area to be settled and by their large 
initial settling density (0.5 to 1.14 young colonies for each 
square meter of free coral rock area). This initial advantage and 
their more rapid growth rate causes them to form a varied coral 
landscape within a reltively short period of time with diversified 



13 



hiding places and food sources providing niches for a reef 
biocenosis rich in species and numbers. 

Thus, it is predicted (as in comparable structures in Red Sea 
coral reefs) by the indicator species present, that with further 
undisturbed development, an Acropora humilis -hyacinthus-assemblage 
or an Acropora - Pocillopora eydouxi - community will develop into the 
future biophysiographic zone of the reef edge and the neighboring 
reef flat. 

Acknowledgements 

I gratefully acknowledge financial assistance received from 
Deutsche Forschungsgemeinschaf t. Thanks are also extended to Dr. 
P.J. Beveridge, University of South Pacific, Suva/Fiji, for his 
willingness to discuss my observations and results, to Roger Moffet 
(Directorate of Overseas Surveys, U.K.), Sam Rawlings (Fisheries 
Officer), Colin Restston (District Commissioner) and Graham 
Worthington (Gilbert & Ellice Islands Development Authority), all 
of these on Funafuti in this period, for their help, to Fried 
Theissen, Bochum, for carefully preparing the illustrations, and to 
Mrs. Sheila Ives, Bochum, for correcting this manuscript. 



14 



References 

Aerial Photos of Ellice Islands. Film I, 11th of July, 1971, 

No. 73/74 (18,000 ft.): 137/206, 159/160, 178/179, 
188/189 (all: 5,000 ft., scale 1:10.000). Directorate of 
Overseas Surveys, Surrey, Great Britain. 

Agassiz, A. (1903): The coral reefs of the tropical Pacific. 

Memoirs Mus. Comp. Zool. at Harvard College, ^: Cambridge, 
U.S.A. 

Baines, G.B.K., P.J. Beveridge and J.E. Maragos (1974): Storms 

and island building at Funafuti Atoll, Ellice Islands, Proc. 
Second Int. Coral Reef Symp. 2, Great Barrier Reef 
Committee, Brisbane: 485-496. 

Blumenstock, D.I. (1958): Typhoon effects upon Jaluit Atoll 

in the Marshall Islands. Nature (London), 182: 1267-1269. 

Blumenstock, D.I. (1961): A report on typhoon effects upon Jaluit 
Atoll. Atoll Res. Bull. 75j 1-105. 

Blumenstock, D.I., F.R. Fosberg, and C.G. Johnson (1961): The 

re-survey of typhoon effects on Jaluit Atoll in the Marshall 
Islands. Nature (London), 189 : 618-620. 

David, T.W.E. and G. Sweet (1904): The geology of Funafuti. 

In, The Atoll of Funafuti : 61-88, London. Royal Society. 

Finckh, A.E. (1904): Biology of the reef-forming organisms at 
Funafuti Atoll. In, The Atoll of Funafuti : 125-150, 
London. Royal Society. 

Fishelson, L. (1973): Ecology of coral reefs in the Gulf of Aqaba 
(Red Sea) influenced by pollution. Oecologia (Berl.) 12: 
55-67. 

Loya, Y. (1976a): Recolonization of Red Sea corals affected 
by natural catastrophes and man-made perturbations. 
Ecology 57(2) : 278-289. 

Loya, Y. (1976b): The Red Sea coral Stylophora pistillate 
is an r strategist. Nature (London), 259 : 478-479. 

McKee, E.D. (1959): Storm sediments on a Pacific atoll. J. 
Sedimentary Petrology 29^: 354-364. 

Maragos, J.E., G.B.K. Baines, and P.J. Beveridge (1973): 
Tropical cyclone creates a new land formation on 
Funafuti Atoll. Science, 181: 1161-1164. 



15 



Mergner, H. (1979): Quantitative okologische Analyse eines Riff- 
lagunenareals bei Aqaba (Golf von Aqaba, Rotes Meer) . 
Helgolander wiss. Meeresunters. 32: 476-507. 

Mergner, H. (1981): Man-made influences on and natural 

changes in the settlement of the Aqaba reefs (Red Sea). 
Proc. Fourth Int. Coral Reef. Symp., Manila. _1: 
193-207. 

Mergner, H. and H. Schuhmacher (1981): Quantitative Analyse 
der Korallenbesiedlung eines Vorrif fareals bei Aqaba 
(Rotes Meer). Helgolander wiss. Meeresunters. 34: 
337-354. 

Pearson, R.G. (1981): Recovery and Recolonization of coral 
reefs. Marine Ecology. Prog. Ser. 4_ (1): 105-122. 

Schuhmacher, H. (1977): Initial phases in reef development, 
studied at artificial reef types off Eilat (Red Sea). 
Helgolander wiss. Meeresunters. 3i0_ : 40-411. 

Stoddart, D.R. (1963): Effects of Hurricane Hattie on the 
British Honduras reefs and cays, October 30-31, 1961. 
Atoll Res. Bull. 95: 1-142. 

Stoddart, D.R. (1965): Re-survey of hurricane effects on the 
British Honduras Reefs and Cays. Nature (London), 207: 
589-592. 

Stoddart, D.R. (1974): Post-hurricane changes on the British 
Honduras reefs: Re-survey of 1972. Proc. Second Int. 
Coral Reef Symp. 2, Great Barrier Reef Committee, 
Brisbane: 473-483. 

Tunnicliffe, V. (1981): Breakage and propagation of the stony 
coral Acropora cervicornis . Proc. Natl. Acad. Sci. USA, 
74 (4): 2427-2431. 

Woodley, J.D. et al. (1981): Hurricane Allen's Impact on 
Jamaican Coral Reefs. Science, 214 : 749-755. 



16 



Fig. 2. Approach flight to Funafuti atoll. Aerial photograph 
from south. Location of sections 6 and 7 below top of 
propeller. 

Fig. 3. Tengako Island, lagoon-side. View from SSE to Amatuku 
Isle (middle background) and the location of reef sec- 
tion 2 (middle foreground). 

Fig. 4. Funafuti Island, lagoon-side. The destroyed reef edge 
of reef section 3, 120 m south from the jetty of 
Fongafale, covered by thick layers of Dictyota bartaysii. 

Fig. 5. Fuafatu Island, lagoon-side. Inner reef at the NE-end 
with reef section 4. Parts of the Montipora foliosa - 
zone (middle) with Acropora hyacinthus (above left), 
Goniastrea sp.? (left) and Pocillopora eydouxi (below 
right). 

Fig. 6. Fuafatu Island, reef section 4. Reef flat with the 
well-developed Pocillopora eydouxi-zone. 

Fig. 7. Fuafatu Island, reef section 4. Crevasses in the reef 

flat with the Acropora hyacinthus - Pocillopora eydouxi - 










jjlfc#<ar ' ; -/ jfe» L^j' 


& ; f^H 




■Br r.'J 




>• 




J 


[ Y? % 7 


1 1 - 


V-' 4 


Si 


■&'-** 




■ 


• 




■ 




\ 


7 

00 




19 



Fig. 8. Fuafatu Island, SE-side. Outer reef with reef section 
5. Former living reef platform with extended Montipora 
foliosa- crusts (middle) and dense Hal imeda- stocks (above, 
right). 

Fig. 9. Funafuti atoll, south-end with the location of reef 
sections 6 and 7 (see arrows) near Tutanga Island. 
Aerial photograph from west. 

Fig. 10. Tutanga Island, outer reef at the SW-side. Boulder zone 
of reef section 6 with large boulders thrown by cyclones 
from western directions onto the reef flat. View from 
WNW. 

Fig. 11. Tutanga Island, inner reef at the SE-side with location 
of reef section 7 (middle right) . View from NW. 




Monttpota 



coral sand beao> rtxk Ponlhpors Acropora Faviidae 

* debris eydovxi hyaonthut Poritei 

* cerymbosa Heuopora dichotomy 

Funafuti Atoll 

Diagrams ( cross sections) of the 7 reef sites investigated flQ 12 
scales, horizontal r : woo vertical 1. soo J 



aio.at 




©03 Goniasfrea (1 ^diameter in cm) Funafuti AtOll 

• Acropora n— n , , Recolonization of the reef platform in section i 

^ *<•* coral rock 

m Pocillopora on the north end of Tengaho, 

/g. _, . I v.l coral sand ; n J u ly 7973 8h months after hurricane ' Bebe 

©P'^ygyra ^ 4 

% Pontes \^2 coral debris e ' . ,_ 5 depth in mefers 



fig. 13 




Caulerpa racemosa occidentalis 
Halimeda macroloba 
Halimeda opuntia 
Udothea orientalis 
Dictyota bartaysii (?) 
A cropora fs = diameter in cm) 
Pocillopora damicornis 
Pontes lutea 
Faviidae 
coral rock 
coral sand 
coral debris 



Funafuti Atoll 

Recolonization of the reef platform in section 3 near Fongafale, 
in July, 1973, BYi months after hurricane 'Bebe' 



Scale 



5 m 



• 1,z depth in meters 



fig. I 4* 



Table 1: Zonation and recolonization of reef section 1 



Reef zone; length 
and depth in m 



Structures and 
biophysiographic zones 



Flora and fauna (selections- 
coral recolonization 



Tide zone and 
shore channel: 
length 15 - 25 m, 
at 25 m, depth 1m 

Abraded reef flat: 
length 110-150 m, 
depth 1 - 2 m; 

at 25 -50 m: 

at 70 m : 

at 75 - 125 m: 

at 100- 150 m: 

Former living 
reef platform: 
length 10 -25 m, 
depth 2-3 m ; 
at 135 - 160 m: 



Reef edge and 

reef slope: 

at 160 m, depth 4 m 



Upper fore reef: 
from 160- 180 m 
outwards, 
depth 4- 6 m 



Coral sand and coral debris 
on greatly cleft beach rock: 
Ectocarpaceae- zone 



Sparsely jointed 
abrasion zone with mud, 
sand and debris: 

Chlorodesmis - zone 
Valonia - zone 
Halimeda - zone 
Caulerpa - zone 

Dead coral rock flat 

with coral debris; 

single living massive corals: 

Plesiastrea - zone 



Roughly cleft coral rock 
without preserved 
fine structures : 
Caulerpa - Plesiastrea - zone 

Coral fine sand with plains 
of coral rubble 



Filamentous algal lawn,- 
aside from a few Paguridae 
no visible fauna 



Aside from some species of fish 

no macrobenthic fauna 

and only poor algal growth 

in biophysiographic zones: 

25% Chlorodesmis fastigiata, 

single Valonia ventricosa, 

1% Halimeda opuntia, 

5% Caulerpa racemosa occidentalis 

A few surviving massively growing 

coral colonies in a zone of 25 m 

along the reef edge: 

9 colonies of Plesiastrea (versipora?) 

and 1 of Qoniastrea retiformis, 

each with 0.2 - 0.5 m0,- 

no coral recolonization 

Aside from a few species offish 
no macrobenthic fauna and only 
1 - 5 % Caulerpa racemosa occidentalis; 
no coral recolonization 

Only 7surviving colonies 

of Plesiastrea near the reef slope, 

very few benthic fish species 



Table 2- Zonation and recolonization of reef section 2 



Reef zone; length 
and depth in m 



Structures and 
biophysiographic zones 



Flora and fauna (selection); 
coral recolonization 



Tide zone and 
shore channel : 
length 5 - 10 m, 
depth 0.5 m 

Abraded reef flat: 

length 100 m, 

at 20 -40 m, depth 1-2 m ; 

at 50-90 m: 

Former living 
reef platform: 
length 15 - 20 m, 
depth 1.5 - 3 m 

at 120 - 140 m: 



Reef edge and 

reef slope: 

at 140m, depth 3 -4m 



Coral sand and coral debris 
on greatly cleft beach rock: 
Ectocarpaceae - zone 



Sparsely jointed, 
largely eroded and partly muddy 
abrasion zone with flattened 
surface and small algal stocks: 
Vasum - Caulerpa - zone 

Dead coral rock flat without 
fine structures and mud, 
but roughly divided into blocks 
and channels with debris: 

Acropora humilis - hyacinthus - 
zone 

Tabular Acropora - colonies 
overthrown by storm waves 



Aside from a few Paguridae 
no visible macrobenthic fauna,- 
filamentous algal lawn: 
Ectocarpaceae and others 

Sparse macrobenthic mobile fauna: 
only 1 living Conus species 
and many Vasum turbine/lum,- 
algal growth: Bryopsis pennata and 
Caulerpa racemosa occidentalis 

No surviving coral colonies, 
but numerous resettled 
young colonies of Acropora spp., 
some of Pocillopora, Pontes, 
Ooniastrea and Plarygyra, 
altogether 84 young colonies 
within an area of 136 m 2 

Rich fish fauna, 

locally 1 hydroid species 

along the reef edge 



Upper fore reef: 
from 140m outwards, 
depth 3 - 5 rn 



Coral sand with mud 
and some coral rubble 



More than 50 fish species : 
many Acanthuridae and 
Chaetodontidae, but no Scaridae 



Table 3: Zonation and recolonization of reef section 3 



Reef zone; length 
and depth in m 



Structures and 
biophysiographic zones 



Flora and fauna (selections- 
coral recolonization 



Tide zone, 10 m, 
shore channel: 
length 20 m, 
depth 0.3 m 

Abraded |reef flat: 
length 80 - 100 m, 
depth 1 - 1.5 m 



Former living 
reef platform: 
length 15- 20 m, 
depth 1- 1.8 m 



Coral rock blocks, 

muddy coral sand with rubble: 

Paguridae -zone 



flattened, muddy abrasion zone, 

only partly visible due to 

algal cover, partly with 

coarse rubble and serious 

damages: 

Dictyota bartaysii -zone 

(Fig. 4) 



Largely eroded and flattened, 
slightly muddy coral rock area, 
covered to 75% with Dictyota 



Aside from a few Paguridae 
only sparse macrobenthic fauna; 
almost no macroalgae 

Flourishing algal stocks: 
mainly Dictyota bartaysii, 
additionally Caulerpa racemosa 
occidentalis, Halimeda macroloba, 
Haiimeda opuntia and 
Udothea oriental is ,■ 
herbivorous macrofauna: 
Holothuria leucospilota, 
Acanthuridae and others 

A few gastropods, 
some Holothuria leucospilota, 
many herbivorous fishes : 
Acanthuridae, Chaetodontidae, 
Pomacentridae ( ? ) 



Reef edge and 

reef slope : 

at 120 m, depth 3 m 



Upper fore reef: 
from 120 m outwards, 
depth 3- 3.5 m 



Broken coral rock zone, 

slightly muddy without fine 

structures, 

covered to 50% with Dictyota,- 

overthrown Acropora spp. : 

Acropora - Pocillopora -zone 



Largely muddy coral sand 
with coral debris 



Sparse algal stock; no surviving 
coral colonies, but numerous 
resettled young colonies, usually 
Acropora spp. * Faviid -regenerates, 
altogether G7 young colonies 
within 50 m 2 and 2.5 m breadth 
along the reef edge 

Very rich fish fauna : 
especially Acanthuridae and 
Mullidae, but no Scaridae 



Table 4 : Zonation and re colonization of reef section 4 

Reef zone; length Structures and Flora and fauna (selection); 

and depth in m biophysiographic zones coral recolonization 

Tide zone: ( S) Coral rubble on beach rock, Few visible macrobenthic fauna; 

length 10- 15 m ( N) eroded beach rock plates no macroalgae 

Shore channel : Coral sand with fine No macroalgae, a few fishes; 

length 20 m, coral debris, (S) areas of (S) Acropora formosa - (group) 

depth 0.5 - 1m,- living branched corals 

strong longreef 
current: 15 m/min. 

Reef flat: Abraded coral rock flat, ( S ) dense barrier of 

length 100 m, on it living coral zones: Acropora formosa -(group), 

depth l-i.5(ii (S) Acropora formosa - zone, (N\) Montipora foliosa, 

(M) Montipora foliosa -zone colonies up to 2 m ( Fig. 5), 

(Fig. 5), (N) Pocillopora (N) dense stock of Pocillopora 

eydouxi - zone (Fig. 6) eydouxi ( Fig. 6 ) with some 

P. damicornis, Acropora hum His 
and Millepora dichotoma 

depth 1.5 - 2 m, In the coral rock flat, (S) dense barrier of 

holes up to 4 m deep crevasses up to 10- 15 m long, Acropora hyacinthus ( Fig. 7 ), 

varying in depth, with fine colonies up to 1.5 m0, with less 

coral sand, along the rims A corymbosa and A humilis , 

branched corals, on the (N) Pocillopora eydouxi, 

floor massive stonycorals: in the crevasses Favia, Favites, 

Acropora hyacinthus - Oomastrea, Plesiastrea 

(Pocillopora eydouxi)-zone (versipora ?)and Pontes (lutea *) 
( Fig. 7) 

Reef edge . more than 800 m away ( impossible to reach ) 



Table 5 : Zonation and recolonization of reef section 5 



Reef zone; length 
and depth in m 



Structures and 
biophysiographic zones 



Flora and fauna (selection),- 
coral recolonization 



Tide zone : 
length 10- 20 m 



Coral rubble on beach rock, 
(SW) beach rock plates 



Aside from a few Paguridae 
no visible macrobenthic fauna 



Abraded | reef flat: 
length 50- 80 m 

at 30 m, depth 0.5 m 



length 20- 30 m, 
depth 0.3 - 0.7 m 



Former living 
reef platform: 
length 20-40 m, 
depth 1 - 1.2 m 



Algal ridge: 
length 30m, 
depth 0.8 - 1m 



Abrasion zone: sharp- edged, 
eroded coral rock 

Chlorodesmis - zone 



Shingle zone: coral debris 
encrusted with calcareous 
red algae on largely eroded 
coral rock, few living corals: 
Porolithon - zone 

Rest of the formercoral zone: 
living coral assemblages 
largely reduced by storm 
damages, with overthrown 
Acropora - umbrellas and green 
algal growth: Halimeda- 
Montipora - zone (Fig. 8) 



Chlorodesmis fastigiata 
and Lithothamnion sp. 

Very few Pagurids and 
Gastropods, a few fishes 

Lithothamnion sp., Porolithon sp. 

Porites lutea (?), 

some Gastropods, a few fishes 



Green algae: 4 Halimeda species 
( H. cylindracea, discoidea, macroloba, 
opuntia) and Udothea orientalis; 
surviving corals: Acropora humilis 
and hyacinthus, Goniastrea, Fa via, 
Montipora foliosa -crusts (Fig. 8), 
Plesiastrea (versipora ?), 
Pocillopora damicornis + P. eydouxi, 
Millepora sp./ recolonization: 
a few young Acropora - colonies 



Typical algal ridge: minimal Aside from Lithothamnion and 

damage, coral rubble cemented Porolithon no macroalgae; 
with calcareous red algae no living corals, a few fishes 



Reef edge with the surf zone and the reef slope with its spurs and grooves 
110 - 200 m away were not possible to reach due to the breakers 



Table 6 : Zonation and recolonization of reef section 6 



Reef zone/ length 
and depth in m 



Structures and 
biophysiographic zones 



Flora and fauna (selection), 
coral recolonization 



Tide zone ■■ 
length 10- 15 m 

Abraded reef flat, 
length 100- 150 m, 
ai 20 m, depth 0.5 ■ 1 m 



at 30 m, depth 08- 1 5 m 



at 50 -80 m, depth 1m 

rapid surface current 
( 40cm/s) outwards 



Coral rubble, but no beach 
rock 

Abrasion zone : 
flattened, sharp- edged 
eroded coral rock plain 
without fine sediment 

Boulder zone (Fig. 10): 
boulders of up to 2 m 
thrown by the storm onto 
the coral rock 

Algal ridge coral debris 
cemented with calcareous 
red algae: 
Porolithon- zone 



Aside from a few Paguridae 
no visible macrobenthic fauna 

No macroalgae,- only a few 
Gastropods, crabs and fishes 



On the coral blocks only 
blue algae and Gastropods, 
single fishes beneath the blocks 

No macroalgae, only Porolithon sp. 
and Lithothamnion sp.; 
aside from fishes 
no visible macrofauna 



Reef edge at loom 
and upper fore reef: 
length 30 m, 
depth 3-5m ; 
strong surface current 
(30cm/s)outwards 

at depth 3-5 m: 

at depth 5 - 10 m. 



Reef edge largely cleft by 
surf erosion and settled 
by branched corals,- 
spurs and groove s 
lead to deep canyons 
with plentiful live : 

Pocillopora - Millepora 
platyphylla - zone, 
Acropora corymbosa - 
A. hum Hi s - zone, 
Millepora dichotoma - zone 



Outwards increasing numbers of 

coral coloni es : 

2 species of Pocillopora 

and plenty of Millepora platyphylla, 

then numerous Acropora humiiis 

and A corymbosa, 

single A.hyacinthus and dense 

barriers of Millepora dichotoma ; 

fish fauna rich in species + numbers 



Table 7- Zonation and recolonization of reef section 7 

Reef zone: length Structures and Flora and fa una ( selection); 

and depth in m biophysiographic zones coral recolonization 



Tide zone : 
length 10 m 

Abraded reef flat: 
length 100-i30m, 
at 30 m, depth lm 



No beach rock, coarse Aside from a few Paguridae nearly 

coral rubble with sand areas no visible macrobenthic fauna 



Abrasion zone: 
flattened, sharp -edged, 
muddy coral rock 



No macroalgae,- 

only a few macrobenthic 

faunal species and fishes 



at 40 m, depth l-2m 



at 70 m, depth 2-3 m- 
strong surface current 
( 25 cm/s) outwards 



Living reef platform; 
length 50- 100 m, 
depth 3- 5 m; 

at 100m, depth 3m, 

at 125 m, depth 3 -4 m, 

at150-200m, depth 4 -5 m 

Reef edge: 
at 170- 200 m; 
only 5 - 8 m 
of visibility 



Microatoll zone: 

numerous microatolls cover 

10-30%: Porites lutea -zone 

Single plains of coral debris 
between bigger microatolls-- 
Porites - Udothea - zone 



Pillar zone: coral rock plain 
with big microatolls, 
towards the southeast 
increasingly divided into 
pillars densely settled by 
living corals; 
between them sand floor 
with mud and fine debris: 
Heliopora - Porites - zone, 

Pocillopora - Miilepora 
dichotoma - zone, 
Pocillopora - Acropora 
hyacinthus - zone 



Microatolls of Porites lutea (?) 
up to an heigth of 0.2 m, 
many of them dead and muddy 

Aside from Porites lutea 
Udothea orien talis and 
Liihothamnion, additionally single 
Miilepora dichotoma and 
Acropora hyacinthus,- a few fishes 

Microatolls of Heliopora coerulea 
with up to 2.5 m0 and of 
Porites lutea with up to 5 m 

Pillars with Pocillopora eydouxi 
and Miilepora dichotoma 

Pillars with Pocillopora and some 
Acropora species (corymbosa, 
formosa, humilis and hyacinthus, 
the latter with umbrellas 
of upto 3m0) 

Some Conus- species and Tridacna, 
only a few fish species 



ATOLL RESEARCH BULLETIN 
No. 285 



STATUS AND ECOLOGY OF MARINE TURTLES 
AT JOHNSTON ATOLL 



By 
George H. Balazs 



Issued By 

THE SMITHSONIAN INSTITUTION 
Washington, D- C, U -S • A* 
May 1985 



STATUS AND ECOLOGY OF MARINE TURTLES 
AT JOHNSTON ATOLL 

By 
George H. Balazs 1 

INTRODUCTION 

The aim of this paper is to consolidate all available information on 
marine turtles at Johnston Atoll, and to present the results of a short- 
term tagging study recently conducted there. The importance of this work 
rests on the fact that it has never been done before, that marine turtles 
are listed under the U.S. Endangered Species Act (since 1978), and the 
atoll is a National Wildlife Refuge. The Defense Nuclear Agency has 
operational control of Johnston with the primary mission of maintaining 
nuclear readiness for the resumption of atmospheric testing, should it be 
so directed. Several other organizations are present under Defense 
Nuclear Agency stewardship, including an Army chemical storage facility, 
a Coast Guard loran station, a NOAA weather station, and a civilian 
support contractor, Holmes and Narver, Inc. The U.S. Fish and Wildlife 
Service in Honolulu cooperatively manages the area as part of the 
National Wildlife Refuge System. 

Johnston Atoll is located at lat. 16°45'N, long. 169°31*W, and is 
one of the most isolated atolls in the world. The land area consists of 
four islands (Johnston, Sand, Akau, and Hikina) totaling only about 2.8 
km 2 , most of which is man-made. The surrounding reef covers an area 11 
by 22 km. Johnston is one of the best studied atolls in the central 
Pacific, due to its small size and extensive use for military purposes 
over the past 45 years. A comprehensive summary of the atoll's natural 
history, including all known scientific studies up to 1973, has been 
compiled by Amerson and Shelton (1976). The ecological significance of 
the atoll is described separately in this publication (p. 361-368) by 
four prominent ecologists. 

Much of the previous research conducted at Johnston has been on the 
terrestrial fauna and flora, with major emphasis on seabirds. The 
studies on the marine environment and biota have been mainly centered in 
the lagoon. Virtually no work has been done off the south shore of 
Johnston Island. This has been due in part to safety and security 
restrictions, and poor diving conditions. Since most of the turtles 
found at the atoll occur in this region, it is perhaps not surprising 
that they have received so little attention over the years. 



1 Southwest Fisheries Center Honolulu Laboratory, National Marine 
Fisheries Service, NOAA, Honolulu, Hawaii 96812 



The Army plans to construct a large-scale incineration facility on 
Johnston Island to destroy chemical agents and munitions stored there. 
The storage bunkers are along the south shore of the island adjacent to 
West Peninsula, the site where the plant will be constructed. Compre- 
hensive information on this project, Johnston Atoll Chemical Agent Dis- 
posal System (JACADS), has been presented (U.S. Army Corps of Engineers 
1983). Construction of the facility is planned to begin in late-1985, 
and take 3 years to complete. The number of personnel on the island will 
double from the present 350. Factors that have not yet been decided, or 
are classified for security reasons at present, include the life of the 
facility, the total number and kinds of munitions to be incinerated, and 
the disposition of certain nontoxic byproducts. Initially, at least 
72,000 rockets containing 345 metric tons of nerve agent (GB and VX) are 
scheduled to be processed. 

This paper is the culmination of work commissioned by the U.S. Army 
Corps of Engineers, Pacific Ocean Division, to obtain baseline data on 
marine turtles at Johnston Atoll. The study was prompted by the absence 
of information on these reptiles, their protected status under Federal 
law, and the proximity of the JACADS project. Recommendations given in 
this paper will help ensure the conservation of Johnston's turtles, as 
requested in the terms of reference for the study. 

HISTORICAL OVERVIEW OF TURTLES 

There are few accounts of sea turtles at Johnston Atoll in the 
literature. Amerson and Shelton (1976) summarized, as follows, all 
information known to them as of late 1973 (p. 112): 

"Reptiles, Species Accounts - There are no general references that 
illustrate the reptiles of Johnston Atoll. Taxonomy of the turtles 
follows Carr (1972) and Amerson (1971). 

"BLACK SEA TURTLE Chelonia agassizi 

" Status - Regular uncommon visitor; known from the lagoon, offshore 
Johnston Island, and Sand Island. 

" Observations - Brooke (MS.), who visited Johnston Atoll in March 
1859, commented about the lack of turtles: "The reefs are covered 
with fish of various kinds. Mullet abound, but there are no turtles.* 
Wetmore (MS. a and b) likewise, recorded no turtles at Johnston Atoll 
in July 1923. 

"POBSP [Pacific Ocean Biological Survey Program] personnel recorded 
sea turtles in the shallow marginal reef area west of Johnston Island 
in July 1963. An adult (USNM 163581) was collected 20 November 1966 
on the beach of Sand Island. Island personnel in 1973 reported 
seeing 10 to 12 turtles offshore of Johnston Island throughout the 
year. A longtime resident estimated harvesting 12 to 15 per year. 

" Annual Cycle - The Black Sea Turtle apparently visits Johnston Atoll 
year-round. No records exist of it breeding on the atoll, although 



perhaps it did in small numbers prior to inhabitation by man. This 
species breeds during the summer in the northwestern Hawaiian Islands, 
especially at French Frigate Shoals (Amerson, 1971)." 

The validity of the species account by Amerson and Shelton (1976) 
will be a subject of discussion later in this report. 

In December 1892, Captain John Cameron of the schooner Ebon stayed at 
Johnston Atoll for a month after sailing directly from Laysan Island in the 
Northwestern Hawaiian Islands. The account of this visit (Farrell 1928) 
mentions sea turtles, but was not cited by Amerson and Shelton (1976). The 
relevant sections of Farrell (1928) are as follows (p. 402-405): 

"Our first call was at Johnston or Cornwallis Island, five hundred 
and sixty miles south of Laysan and southwest of the eight islands 
of Hawaii proper. We found a good berth in its lagoon, and in a 
pretty little cove, on a beach of white sand, was an ideal spot for 
our tent. Near by were ruins of shanties built years before by a 
guano company; there also was a well, with pumps and pipes intact, 
which we cleaned and put in order. 

"Signs of men, signs of shipwreck! We stumbled across two boats, 
both hauled above high water, one in fair condition, the other badly 
smashed; and in the craft were harpoons and lances and some bird 
shot in bags. The condition of the better boat, which was well 
worth the repairs I decided to give it, indicated that the men who 
left it there had been rescued. Else why should they have abandoned 
a tolerably seaworthy craft on a desert island? 

"Our catches of sharks at Johnston were only fair, because our bait 
was principally sea birds, which the brutes did not relish as they 
had the flesh of hair seals; but our hauls almost filled our con- 
tainers with liver oil. Now and then we took things more easily: 
'Spell 0!' was passed, and we hunted turtles. One of the men employed 
the Rusaie method of taking them by anchoring a few captive females 
near the beach to attract the bulls. It succeeded admirably and 
helped us greatly with attractive bait for shark fishing. 

"We were standing to sea, bound to Fanning Island, when from the mate, 
who was at the masthead, came a cry of 'Sail 0!' A bark under full 
sail was heading for us. Through the telescope we could see that she 
was a whaler: that was made evident by boats hanging from her davits 
ready for immediate use. I lost no time in pulling to her with some 
turtles and two pigs, welcome additions to the fare of a vessel long 
at sea and especially for Christmas dinner, as the day was December 24." 

Votaw (1943) included some of Captain Cameron's comments about tur- 
tles in a short historical paper on Johnston Atoll. Turtles were also 
mentioned, but again not cited by Amerson and Shelton (1976), in a 
Honolulu newspaper article by Benson (1953). Describing the dynamiting 
necessary from 1939 to 1942 to clear out coral heads in the lagoon, Benson 
(1953) stated that: 



"Interest was added to the process by the presence of numerous huge 
sharks. There was one monster in particular who demonstrated his 
prowess one day by swallowing a sea turtle - whole - at one gulp." 

In recent years, sea turtles at Johnston Atoll have been discussed by 
Balazs (1978, 1980b, 1980c, 1982d). Two of these papers (Balazs 1980b, 
1982d) made recommendations that stressed gathering baseline data on this 
little-known and isolated turtle population. A brochure describing 
National Wildlife Refuges in the Pacific briefly mentions that the green 
sea turtle is among the marine life found at Johnston (U.S. Fish and 
Wildlife Service, MS.). 

Except for the single specimen in the U.S. National Museum listed by 
Amerson and Shelton (1976), there is no indication of scientific personnel 
having examined, tagged, or studied sea turtles at Johnston Atoll. Start- 
ing in 1978, turtle sighting report forms have been sent to resident per- 
sonnel at the atoll, but only limited information could be acquired by 
this method (Balazs 1982d). Casual observations and counts of turtles 
from shore have recently been recorded in trip reports by Ludwig (1982), 
Ludwig et al. (1982), and Nitta (1982). Applied Eco-Tech Services (1983) 
included a discussion of turtle sightings in their consultancy report on 
water quality. 

ASSESSMENT METHODS 

Two field studies, totaling 28 days, were conducted at Johnston Atoll 
to accomplish the assessment. The first phase of study was September 29- 
October 13, 1983 and involved two workers. The second phase was November 
3-17, 1983 involving five workers. In addition, a preliminary 2-day 
planning visit was made by one worker on August 30-September 1, 1983. 

Capture Efforts 

Efforts to capture turtles alive and unharmed were undertaken using 
large-mesh tangle nets, scoop nets, and scuba to facilitate capture by 
hand. All three of these methods have been successfully employed to study 
green turtles in coastal waters of the Hawaiian Islands (Balazs 1976, 
1982b). 

The tangle nets were made of 2-mm diameter nylon line, with a 
stretched mesh of 46 cm (23 cm square mesh), and a depth of 3.5 m. The 
length of the nets ranged from 9 to 40 m. The nets were set at the sur- 
face extending vertically through the water column. They were deployed 
and retrieved close to shore using a small boat at sites recommended by 
resident personnel, or where turtles were seen foraging or in transit. Up 
to five nets were set at one time at different locations. All nets were 
checked from land with binoculars every 1-2 h diurnally to see if a turtle 
had been caught. 

Large scoop nets were used by approaching turtles at the surface with 
a boat. Efforts with scuba were directed at locating and catching turtles 
by hand during the course of underwater surveys. 



Following their capture, turtles were taken ashore for tagging and 
examination for a period requiring up to 2 h. Before being released, 
color photographs were taken to help document morphological features. 

Tagging and Body Measurements 

Turtles were tagged for long-term identification with numbered 
Inconel 2 alloy tags, size 681, custom made by the National Band and Tag 
Company of Newport, Kentucky. Balazs (1980c, 1982a, 1983) describes the 
history of these tags used in Hawaii and their superior corrosion resis- 
tance compared with Monel alloy. The tags measure 25 x 9 x 8 mm, weigh 
3.5 g, and are self-piercing and self-locking when applied with special 
applicators. Depending on the turtle's size, from two to five tags were 
applied to offset tag loss. Tagging sites were the trailing edges of both 
front flippers in the webbing between the third and fourth scales counting 
proximal to distal, in the axillae close to the first scale, and on a hind 
flipper on the inside trailing edge well under the carapace. A secondary 
and potentially long-lasting mark in the edge of the carapace resulted 
from the bone biopsy (described later). 

Short-term visual recognition of tagged turtles after their release 
was made possible by painting a white number on each side of the carapace 
using Dupont Lucite spray paint. Based on studies elsewhere, it was 
estimated that these numbers would remain visible for at least 10 days. 

Observations on the turtles consisted of: straight-line (SCL) and 
curved (CCL) carapace length from the center of the precentral scute to 
the posterior tip of a postcentral scute; straight-line carapace length 
from the center of the precentral to the notch between the postcentrals; 
straight-line and curved carapace width at the widest point (the sixth 
marginal scute); straight-line plastron length along the midline; straight- 
line head width at the widest point; tail length from the posterior rigid 
edge of the plastron to the tip of the tail; straight-line flipper width 
from the claw scale to the sixth scale on the trailing edge; and body 
weight. 

Food Sources and Epizoites 

Food sources were determined by sampling the turtles' stomachs with a 
plastic tube inserted through the esophagus. Small amounts of water were 
introduced and aspirated to obtain food material. In addition, unswal- 
lowed particles of food were removed from the mouth, and fecal material 
that could be collected was rinsed to isolate incompletely digested food. 
These three field techniques for sampling dietary components are discussed 
in detail in Balazs (1980a). Observations made of turtles feeding at 
specific sites also permitted the direct collection of algal forage during 
underwater surveys. 



2 Reference to trade names does not imply endorsement by the National 
Marine Fisheries Service, NOAA. 



Food items were preserved in dilute Formalin and identified to the 
lowest taxon possible. Frozen bulk samples collected from the foraging 
habitat were biochemically analyzed to determine major nutrients and 
mineral composition. 

Epizoites found on the skin and hard parts of the turtles were 
scraped off, preserved in dilute Formalin, and identified to the lowest 
taxon possible. 

Biopsies and Blood Sampling 

Biopsies of bone and lamina were taken with a saw by cutting a small 
triangular piece from the edge of the 10th left marginal scute. Depot fat 
was sampled from directly under the skin by making a 2-3 cm incision in 
the inguinal region. Tissue sampling procedures described by Rainey (1981) 
and Menzies et al. (1983) were used as a guide for this work. Bone and fat 
samples were frozen in glass vials and stored for future analyses of radio- 
nuclides and heavy metals. 

Blood collection followed the methods described by Owens and Ruiz 
(1980) and Bentley and Dunbar-Cooper (1980). A needle and syringe were 
used to draw blood from the paravertebral sinus on either side of the 
midline of the dorsal neck surface. The blood was centrifuged and sepa- 
rated into packed cells and serum. Sera were frozen and analyzed for 
testosterone levels to determine sex. Packed cells were refrigerated and 
analyzed within a few hours for cholinesterase activity. The 17-Minute 
Manual Method, routinely used at the Johnston Island medical facility to 
detect anticholinesterase intoxication in humans, was used for analysis of 
turtle blood. 

Underwater Surveys 

Underwater scuba surveys were made to census turtles, locate and 
assess prominent foraging and resting habitat, and gather other ecological 
data. Two or three divers working together within visual range carried 
out the surveys. All surveys took place during the daytime. 

Terrestrial Surveys 

Terrestrial surveys were conducted along the coastlines of all four 
islands for the purpose of locating possible nesting and basking habitat. 
Systematic observations from shore were also made of coastal waters. 

Personal Interviews 

To compile anecdotal information, fishermen and divers, especially 
those who have been at Johnston for many years, were interviewed. Requests 
were also made to examine photographs in private collections showing 
turtles caught during past years. 



Literature Search 

The published and unpublished literature pertinent to Johnston Atoll 
was reviewed. This search included articles in the two major Honolulu 
newspapers. All known historical reference to turtles at the atoll have 
been presented in the previous section of this report. However, the lit- 
erature review also encompassed articles on perturbations to the environ- 
ment that could be of significance to turtles or their habitat. 

An inquiry was made to the U.S. National Museum (Washington, D.C.) to 
obtain further data on the specimen mentioned by Amerson and Shelton 
(1976) as having been collected in 1966 "...on the beach of Sand Island." 

FINDINGS 

Results of Capture Efforts 

A total of 21 turtles were captured (Table 1). All were green tur- 
tle, Chelonia my das , taken with nets, and no repeat captures were made. 
There were no scars indicative of old tags being shed. No turtles were 
caught with scoop nets or by hand, due mostly to turbid water conditions 
and the rapid diving behavior of the turtles when approached by boat. The 
locations selected for the nets were exclusively off the south shore of 
Johnston Island (Fig. 1). Nearly all of the turtles sighted during the 
surveys were in this area. The high concentration along this side of the 
island was also confirmed by everyone interviewed. The water off the 
south shore is silt-laden resulting in poor underwater visibility. Rea- 
sonably good water clarity was found at other sites in the atoll. 

The daily netting effort at each location, expressed in meter-hours 
(MH) (length of net by hours fished), is shown in Table 2. During phase 1 
of the field study, nets were regularly set at locations 1-5 and left both 
day and night. However, this sampling procedure proved unworkable because 
of the high incidental capture of eagle ray, Aetobatus narinari , and large 
manta ray, Manta birostris . The entanglement of rays occurred only at 
night or during twilight hours. Once caught, manta rays were able to 
twist with such force that sections of the net became snarled and useless. 
Eagle rays caused less of a problem, but were still able to pull sections 
of floatline underwater and hold them there. Most turtles caught with 
rays under these conditions would have drowned. During the times the nets 
were left out at night, only one turtle was caught, apparently at morning 
twilight (Table 3, tag No. 7451). Although eagle rays were also in the 
net, the turtle escaped injury due to its large size and place of entan- 
glement away from the rays. All netting during phase 2 was conducted 
diurnally (Table 2) to eliminate the problem of accidentally catching 
rays. Shore observations during the daytime indicated that at least some 
turtles avoided the nets. Avoidance would probably not have been possible 
at night. 

Nets were set at 17 sites, but turtles were caught at only 4 of them 
(Table 1). Three of these locations (1, 2, and 3) were close to or 
immediately east of the West Peninsula, and one location (7) was between 
West Peninsula and the southwest corner of the island (Fig. 1). Eight of 



8 

the 21 turtles were caught at location 2; 5 each were caught at locations 
1 and 7; and 3 were caught at location 3. Catch per unit effort was 
considerably better at locations 2 and 7 (589 and 550 MH per turtle, 
respectively; Table 1). A similar catch per unit effort was obtained 
during phase 1 and phase 2 field studies (1,123 and 1,197 MH), although 
twice as many turtles were caught during phase 2 (14 versus 7). 

Species Present 

The turtles captured displayed no clear characteristics that would 
justify their designation as C_. agassizi , the black turtle of the east 
Pacific. The large size of the adults, the contour of the carapace, and 
the color of the plastron were all mostly consistent with C_. mydas . An 
exception was a specimen that had a strongly tapered posterior to the 
carapace, and a moderate amount of gray pigment in the ventral surface of 
the marginal and postcentral scutes (tag No. 7551). These features are 
pronounced in C_. agassizi , including dark pigment throughout much of the 
plastron. This turtle therefore seems to be intermediate between agas- 
sizi and mydas . None of the other turtles captured, nor those remembered 
by Cris Balubar, a resident employee and former turtle fisherman, had dark 
pigment in the plastron. 

Carapace color and pattern varied considerably, ranging from predomi- 
nantly tan with brown radiations (tag No. 7451) to olive with black flecks 
(tag No. 7517). When seen in the water before capture, the carapace of 
most turtles at Johnston is masked by a layer of silt. Carapace colora- 
tion within other green turtle populations, such a6 in Hawaii, is known to 
vary with stage of maturity, sex, and possibly even environmental factors 
(Balazs 1980c). However, the carapace and dorsal skin surface of adult C_. 
agassizi is always predominantly black. 

The observation by Amerson and Shelton (1976) that the black turtle 
occurs at Johnston is invalid based on findings of this present study. 
Amerson and Shelton' s (1976) nomenclature appears to have been founded 
almost completely on their citation of Carr (1972), who stated (p. 25): 

"...The black turtle of the eastern Pacific lacks the numbers to 
withstand that abuse, and may well become an incidental casualty 
along the American mainland shores. To my eye, however, the black 
turtle stock occurs elsewhere — in the Galapagos Islands, among the 
mid-Pacific Islands, and in parts of the Indian Ocean. With its 
range extending through so much territory, the complete loss of the 
Mexican and Central American colonies might not obliterate Chelonia 
agassizi ; but here again, the name, as I am using it, surely covers a 
number of hitherto unnamed races. The sooner these are properly 
defined, the sooner concern over their plight will be generated." 

No other species of sea turtle was seen during the field studies. An 
unverified sighting of a hawksbill turtle, Eretmochelys imbricata , is 
listed by the U.S. Army Corps of Engineers (1983). In addition, four 
turtles, thought to be hawksbills, were reported in shallow water off the 
northeast corner of Johnston Island in September 1980 (R. J. Novak in 



litt. to G. H. Balazs). Cris Balubar and others interviewed indicated 
that only green turtles have ever been seen by them within the atoll. 

The leatherback turtle, Dermochelys coriacea , has been observed on 
several occasions by personnel trolling for fish outside the atoll. Cris 
Balubar saw one about 11 km to the north of the atoll. In 1981, a large 
decapitated (but still moving) leatherback was seen, apparently after 
being accidentally hit by a boat. Efforts to gaff the turtle and bring it 
aboard proved unsuccessful. 

Population Structure 

Body measurements and weights presented in Tables 3, 4, and 5 indi- 
cate that 60% (14) of the turtles captured were mature adults. Turtles 
<82.9 cm SCL were estimated to be immature (see Balazs 1980c for a discus- 
sion of size categories). The proportion of adults in the Johnston popu- 
lation is therefore substantially greater than at coastal areas studied in 
Hawaii. At a comparable foraging area on Molokai, Hawaii only 9% of 81 
green turtles sampled with nets were adults. The sighting of turtles 
during surveys at Johnston, along with information resulting from inter- 
views, confirmed that the population is composed of mostly large turtles. 
The smallest turtle captured was 57.4 cm SCL; however, a few others esti- 
mated to range down to 35 cm SCL were seen during the surveys. 

The age structure and growth rates of turtles at Johnston are pres- 
ently unknown. In the Hawaiian Islands, green turtles are estimated to 
take 11-59 years to grow to an adult. Growth rates have been found to 
differ significantly among resident foraging areas within the archipelago 
(Balazs 1982b). The high percentage of adults at Johnston could be caused 
by several factors, including rapid growth rates, low recruitment of small 
turtles to the population, high predation and mortality of small turtles, 
and low predation and mortality of adults. 

The 15 turtles that were weighed ranged from 63.6 to 151.4 kg 
(Table 5). The mean weight of three adult males was 104.8 kg (range 84.5- 
115.9 kg). The mean weight of six adult females was 112.0 kg (range 87.7- 
115.4 kg). The largest turtle ever caught by Cris Balubar was a 186 kg 
male. 

Testosterone levels were determined from blood samples of 12 turtles 
for sex determination. The sex of six other turtles, all adults, could be 
determined by external features (i.e., a long, large tail for males). 
The sex ratio of this 18-turtle sample was 2.6:1 in favor of females 
(Table 6). If only immature turtles are considered, the ratio was still 
3:1 in favor of females. Of special interest among the immature turtles 
was a fairly large specimen (79.1 cm SCL, 74.1 kg) that still had a short 
tail (18.2 cm). The testosterone level showed this turtle to be a male. 

Abundance and Distribution 

Capture and tagging efforts were focused at principal aggregation 
sites of turtles. Seven turtles were tagged during phase 1 from this 
area. After an interval of 20 days, phase 2 capture efforts yielded 14 



10 

more turtles, none of which had been tagged earlier. Because no recap- 
tures were made, these data alone do not permit an estimate of the number 
of turtles present along Johnston Island's south shore. 

Only one paint-marked turtle was resighted. This turtle was observed 
at the surface near dive location I (Fig. 2) by a resident employee sail- 
ing a Hobie cat. The turtle dove vigorously when approached. Turtles 
have been occasionally seen in this general area, which consists of a 
dredged turning basin and ship channel along the eastern portion of John- 
ston Island's north shore. This turtle (tag No. 7485), had been captured 
6 days earlier on November 6 at net location 2, and released shortly 
thereafter at the port facility on the north shore (Fig. 1). All 7 
turtles captured during phase 1 and 12 caught during phase 2, were 
released at this site. The other two turtles (tag No. 7560 and 7565) 
from phase 2 were transported by truck and released at the south shore. 
The short distance (2 km) around the island, in relatively calm water 
between the north and south shores, should not have presented an obstacle 
to the turtles. Biotelemetry has shown that immature green turtle have a 
well-developed homing ability on their resident habitat (Ireland 1979a, 
1979b). Furthermore, adults can find their way across hundreds of 
kilometers of ocean when migrating between resident areas and nesting 
beaches (Hirth 1971; Carr 1972). 

Only three turtles were seen during 26 diving surveys with scuba 
totaling over 22 h, or 46 man-h, of bottom time (Fig. 2). All three 
turtles were swimming when sighted. While it was not possible to approach 
close enough to capture them, the turtles nevertheless did not flee in the 
manner seen when encountered by boat at the surface. The turtles were 
sighted at dive locations D, J, and P off the south shore in the same 
general area where nets were set. Most of the dive surveys (18 of 26) 
were made here, but poor underwater visibility, ranging from at best 10 m 
to as low as 1.5 m, greatly limited the possibility of seeing swimming 
turtles. However, careful systematic searches of the bottom were made at 
all locations to find places where turtles were sleeping, hence less 
liable to flee from a diver. None was found, although most areas sur- 
veyed, from just several meters offshore out to 1.8 km (dive location 0), 
appeared well suited as sleeping habitat. Sleeping sites repeatedly used 
by green turtles have recognizable marks in the substrate. Except for two 
possible minor sites at dive locations C and P, no habitat was found 
showing such usage. The major sleeping areas for turtles along the south 
shore remain undiscovered, but it seems unlikely they are commuting very 
far. 

As shown in Figure 2, no diving surveys were made to the southwest of 
West Peninsula in the waters downwind of the sewer outfall. In addition 
to raw human sewage, the outfall discharges wash water from the decontami- 
nation procedure used at the chemical storage facility. Other effluent, a 
dense black discharge, was regularly seen from shore during phase 1. This 
was reported to result from flushing of old sewer lines. The discharge 
point of the sewer outfall is immediately southwest of dive location J 
(Fig. 2). No turtles have been seen underwater during the few dive sur- 
veys previously made downwind of the outfall (Ludwig et al. 1982; Applied 
Eco-Tech Services 1983). However, from shore, turtles are commonly seen 



11 

in this area while at the surface. It is possible that the sleeping areas 
are located somewhere along here. Recreational swimming and scuba diving 
are not permitted in waters off the entire south shore of Johnston Island. 

Sightings of individual turtles at the surface, other than along 
Johnston's south shore, were made once at dive location 0, and four times 
in the ship channel at the east end of Johnston Island. In addition, 
three immature turtles were seen over the tops of coral heads in the 
vicinity of dive location X. No turtles were seen in coastal waters by 
observers placed on Akau and Hikina Islands from 0800 to 1700 on November 
10, and at Sand Island from 0800 to 1700 on November 11. No turtles were 
seen during snorkel surveys made by two observers for 80 min on November 9 
in shallow water off the east and north end of Sand Island. No turtles 
were seen by two observers during 30 min of observation on October 12 from 
the abandoned tower near dive location L. These findings, which suggest 
low numbers of turtles and sparse distribution at sites other than John- 
ston's south shore, are consistent with information gathered during inter- 
views. For example, Armyman Tim Snover stated that, over the past 10- 
month period, he had never seen a turtle during six scuba and eight 
snorkel dives in the northern portion of the atoll at Donovan's Reef 
(Fig. 2). Other personnel have seen turtles there, but only occasionally. 

The reports of turtles in abundance along Johnston's south shore, and 
especially off West Peninsula, contrast sharply with the low numbers seen 
elsewhere in the atoll. For example, Ludwig (1982) saw an estimated 30 
turtles during 1 h of observation from West Peninsula at 1800 on Sep- 
tember 15, 1981. Up to five at a time were seen around the tops of indi- 
vidual coral heads. When Ludwig (1982) visited here again on September 
17, he spotted eight turtles during 30 min. Ludwig (1982) also reported 
that personnel frequently visit West Peninsula to watch turtles. During 3 
days in July 1982, Nitta (1982) saw 8-11 turtles while viewing from West 
Peninsula in the late afternoon. Applied Eco-Tech Services (1983) made 
the following comments from surveys conducted along the south shore during 
June 2-11, 1983. 

"Although the survey team was not directing their efforts to turtle 
observations, nearly every head bearing Brvopsis was seen to have one 
to perhaps four specimens of Chelonia mydas in the immediate 
vicinity. .. .Upon sighting the dive boat, these turtles would sound 
rapidly, move away from the area and resurface 20-40 m away. This 
behavior and the uncertainty of the general movements of the turtle 
population during the course of the day precludes an accurate esti- 
mate of the total number of turtles present off the south shore at 
any given time. The algae survey team typically noted 20-25 turtles 
during each morning's efforts (3 h) and a similar number during 
each afternoon. It should also be noted here that this estimate is 
conservative since it represents sightings of surfacing animals only, 
the water clarity being sufficiently poor to prevent sightings of 
submerged organisms." 

At the beginning of phase 1, four to six turtles were commonly 
sighted off West Peninsula, usually within a few hours of high tide. 
However, the number spotted varied considerably (0-13) while motoring 



12 

along the entire south shore at various times throughout the study. These 
counts were undoubtedly influenced by tide stage, changes in visibility 
due to sun angle, and probably a greater awareness by the turtles of an 
approaching boat. At the end of phase 2, on November 16, an observer was 
stationed at West Peninsula for an hour during an incoming high tide. No 
fewer than five and possibly as many as eight turtles were seen; none of 
which appeared to have painted numbers. These data suggest there may be a 
considerable turnover in turtles using the area, and that the total number 
may be many times larger than what can be seen during a several day 
period. 

Food Sources 

Samples of stomach contents, mouth contents, or feces were acquired 
from 13 of the 21 turtles captured. Stomach contents from eight turtles 
contained five kinds of benthic algae, diatoms, filamentous bacteria, 
unidentified fibers, and a single amphipod (Table 7). The green algae, 
Bryopsis pennata var. secunda , was prominent in samples from three turtles 
and Caulerpa racemosa var. uvifera prominent from one turtle. All other 
items were present only in relatively small or trace amounts. Mouth con- 
tents from five turtles contained five kinds of benthic algae (including 
unidentified blue-green algae), diatoms and a single amphipod (Table 8). A 
stomach sample had already been taken from one of these turtles and the 
mouth contents were identical. Of the four turtles sampled only for mouth 
contents, 15. pennata var. secunda was prominent from two £. racemosa var. 
macrophysa prominent from the other two. Identification of fecal contents 
from the single turtle sampled revealed a composition of 75% £. racemosa 
var. uvifera , and 25% B_. pennata var. secunda (Table 8). Differences in 
the digestibility of the two species could have affected these percentages, 
therefore, the actual ratio ingested is unknown. Based on the limited 
sample data presented for stomach, mouth and fecal contents, the size of 
the turtle does not appear to be a factor in the kind of alga eaten. The 
turtles' major food sources ( Bryopsis and two varieties of Caulerpa ) grow 
in prime foraging habitat near the West Peninsula. Turtles were commonly 
seen surfacing and diving in typical foraging behavior over coral heads 
having dense Bryopsis . Interestingly, Bryopsis is not among the 56 known 
species of algae used as food by Hawaiian green turtles, nor has it been 
recorded anywhere else as green turtle forage. However, C_. racemosa is 
eaten at several locations, but seems to be poor forage yielding very slow 
growth rates in green turtles (Balazs 1980c, 1982b). 

Cris Balubar stated that the turtles he recalls cutting open and 
cleaning only contained the two common types of seaweed ( Caulerpa and 
Bryopsis ) found along the south shore of Johnston Island. 

An unidentified cuboidal jellyfish (Cubomedusae) was abundant along 
Johnston's south shore during the early part of pha6e 1. The bell of these 
animals measured about 13 cm long. Many of them settled into the numerous 
depressions between coral heads where torn pieces of Bryopsis also col- 
lected. No evidence was found that green turtles feed on these jellyfish. 
However, other hydrozoans, such as Physalia and Velella , are eaten on an 
opportunistic basis by green turtles in Hawaii (Balazs 1980c). 



13 

Kitchen waste is dumped into the ocean daily at the southwest corner 
of Johnston Island. Pritchard (1982) reports that some green turtles 
scavenge regularly for food scraps discarded in a similar manner at the 
military facility on Kwajalein Atoll. No evidence was found that this 
food source is utilized by turtles at Johnston. 

The circumstances in which a fecal sample was collected from a turtle 
(Table 8) are of interest. This turtle was captured at net location 7 at 
1700 h on November 8 along with another turtle (Table 3). A section of 
the net had become snagged on a nearby coral head, thus preventing it from 
reaching the surface to breathe. The turtle was comatose when recovered. 
Except for contraction of the tail when pulled straight, there were no 
signs of life. Periodic compression of the plastron for 1 h in an 
attempt to ventilate the lungs gave no apparent results. The movement of 
air by this method seemed to only take place through the esophagus, since 
the glottis remained tightly closed. A small diameter plastic tube was 
therefore pushed though the glottis to hold it open and afford passage of 
air. The lungs were then gently ventilated by blowing into them at irreg- 
ular intervals for the next hour with the turtle in a prone position. The 
turtle raised its head and opened its mouth to breathe for the first time 
after being seemingly lifeless for over 2 h. The tube was removed as 
breathing gradually became more frequent, and movement of the flippers 
resumed. The turtle was subsequently left prone overnight to give it more 
time to recover before release. In the morning, 1.5 kg of fecal matter 
were found to have been passed. The turtle swam off and dove in a normal 
manner when released. 

Foraging Habitat 

The principal foraging habitat for turtles at Johnston Atoll consists 
of a narrow band of heterogeneous algal pasture immediately off and along 
the south shore of Johnston Island. To a lesser extent, this feeding zone 
also exists contiguously on the northeast side of the ship channel (dive 
location X, Fig. 2), where Br y ops is alone is present on the tops of coral 
heads. Based on published literature, personal interviews, and surveys 
conducted during the present study, the standing crop densities of benthic 
algae suitable as forage for green turtles are extremely low at most all 
other sites in the atoll. Many kinds of benthic algae occur at Johnston 
(Brock et al. 1966, Buggeln and Tsuda 1966, 1969; summarized by Amerson 
and Shelton 1976). From this literature, Balazs (1982d) previously noted 
that £. racemosa, Codium arabicum , and Gelidium pusillum might serve as 
algal forage for Johnston's turtles, since green turtles elsewhere feed on 
these three species. However, the apparently sparse quantities of the 
latter two species negate any significant benefit that could be derived. 
There are small areas of Caulerpa racemosa var. uvifera in shallow water 
around Sand, Hikina, and also possibly Akau Island that could be used by 
turtles, but no turtles were seen feeding at these locations. 

Four kinds of benthic algae collected during diving surveys along 
Johnston's south shore comprise nearly all of the standing crop that 
exists there (Table 9, dive location J). Three of these, B. pennata 
var. secunda , £. racemosa var. macrophysa , and jC. racemosa var. uvifera , 
were identified as major food sources from the stomach, mouth, and feces 



14 

of turtles. The fourth alga, £. serrulata (Forskal), was commonly seen in 
many areas close to the south shore, often growing in proximity to the two 
varieties of £. racemosa . Since C_. serrulata was not found in any of the 
food samples, the turtles must be actively ignorning this species. Such an 
aversion could be due to metabolites known to be present in some Caulerpa 
that can act as toxic feeding deterrents (Paul and Fenical 1982; Paul 
1983). This deterrence has not been demonstrated in turtles, but the 
subject warrants investigation. Certain algae of the Order Caulerpales are 
food sources for green turtles at a number of locations worldwide. More 
than 100 species are known, and at some atolls in the Pacific very dense 
populations of the algae are present (Meinesz et al. 1981). 

Bryopsis and the two varieties of C_. racemosa sampled fresh 
from foraging habitat off West Peninsula were found to differ considerably 
in nutrient composition (Table 10). Bryopsis contains 2 to 3 times as 
much protein as Caulerpa , and only about 60% the ash content. Bryopsis 
also has 2 to 4 times greater lipid content (ether extract). It should be 
noted that the protein percentages shown in Table 10 were obtained by a 
standard analytical procedure used for terrestrial forage (proximate anal- 
ysis), where total nitrogen is multiplied by a value of 6.25. This may be 
an overestimate for certain marine benthic algae. For example, Dawes and 
Goddard (1978) measured protein in C_. racemosa (variety not stated) from 
Florida by a direct protein extraction technique that gave a content of 
4.8%. Protein content for racemosa in the present study was calculated to 
be 8.0% (for uvifera ) and 9.1% (for macrophysa ) . 

The nutrient composition of Bryopsis collected from two different 
environments is also presented in Table 10. In one, Bryopsis was sampled 
off the top of a coral head where turtles were commonly seen feeding. The 
other collection was made nearby from a depression between coral heads 
where drifts of naturally torn Bryopsis had collected due to low water 
movement. Bryopsis in these depressions is not visible from the surface 
due to high turbidity. Consequently, it is unknown if turtles ever feed on 
this loose material. It was theorized that attached Bryopsis repeatedly 
cropped by turtles might contain higher protein and less fiber (complex 
polysaccharides) due to the constant new growth taking place at the 
grazed ends. Protein levels shown in Table 10 do not support this 
hypothesis, since the loose material is almost 2% higher (25.7% versus 
23.8%) in protein content than the attached alga. However, attached 
Bryopsis is slightly lower in all the fiber components. It is worth 
noting that the drifts of loose Bryopsis can probably remain healthy and 
unattached indefinitely, so long as nutrients are sufficient, and 
currents weak enough, to prevent the drifts from being washed away (D. J. 
Russell, pers. commun. , December 1983). 

The mineral composition determined for the two different collections 
of Bryopsis is very similar (Table 11). The one prominent value for the 
nine minerals measured in these algae is the iron content of C_. racemosa 
var. macrophysa (2,558 ppm). This level is many times higher than that of 
the uvifera variety (90 ppm) or either sample of Bryopsis (88 and 110 ppm). 
No firm explanation can be offered for these data. However, it is possible 
that macrophysa has a high requirement for iron, which therefore may be a 
limiting nutrient to growth. Iron pilings and pipes along the south shore 



15 



of Johnston may supply this nutrient and allow the alga to proliferate as 
it does. This explanation is supported in part by data discussed in 
Russell and Carlson (1978) concerning shipwrecks and the concomitant 
vigorous growth of certain green algae. 

Surveys to characterize and define the habitat limits for the dif- 
ferent algae comprising the turtle foraging zone along the south shore 
resulted in the following findings. The occurrence of attached Bryopsis is 
confined almost entirely to the tops of coral heads that range from not 
more than 2 m beneath the surface, to those that are fully exposed at low 
tide. This environmental range appears to be a necessity for lush Bryopsis 
growth off the south shore. Only a limited number of coral heads fulfill 
the requirement. Many of the Bryopsis covered heads have growth 1-2 m 
down their slope, but thereafter the alga is sparse. A few small patches 
of attached Bryopsis were found on the sides of some heads as deep as 
7.5 m. 

The number of coral heads with Bryopsis was censused off the south 
shore at low tide from a boat on November 16. Between East Peninsula and 
the row of iron pilings (Fig. 1), 30 individual heads were counted, as 
well as a narrow broken ridge in shallow water parallel to shore. Between 
the iron pilings and the southwest side of West Peninsula, 31 heads were 
counted. From West Peninsula to the southwest corner of the island, 18 
were counted. A total of only about 80 heads therefore occur in this 
narrow zone, and most (62) are between East and West Peninsulas. The 
greatest distance of any coral head from shore is about 600 m. Six of the 
heads that were believed to be representative of all the various sizes 
were selected and measured to determine vertical surface area where 
Bryopsis occurs. The heads were irregular and difficult to measure. 
Nevertheless, the areas ranged from approximately 9 to 127 m 2 ; the mean 
was 47 m 2 . An "average" size coral head hosting Bryopsis is therefore 
only about 7 x 7 m. It was estimated that Bryopsis covered 80% of the 
surface area, the remainder being bare coral rock. This surface coverage 
could very well change with season. The distribution of Bryopsis off 
Johnston's south shore as portrayed in a map prepared by Applied Eco-Tech 
Services (1983: Fig. 24) greatly overrepresents the habitat area where 
this alga actually occurs. 

As previously mentioned, there are coral heads to the northeast of 
East Peninsula where environmental conditions are conducive to Bryopsis 
growth and turtles were seen feeding. This area was not as thoroughly 
surveyed. There are probably not more than 15 coral heads there that host 
Bryopsis . 

Turtles foraging on the tops of coral heads with Bryopsis are highly 
visible due to the shallow depth, and the contrasting color of the tur- 
tle's silty-brown shell against the green-black mat of Bryopsis . Because 
turtles are so apparent when foraging at these sites, it was possible to 
ascertain that some heads were being used more heavily than others. Cris 
Balubar also confirmed this point. For many heads, it is essential for 
the tide to be high enough for turtles to swim over the top to forage. At 



16 

a few sites, this condition is never met. Two prominent coral heads off 
West Peninsula, are emergent much of the time, and even at high tide waves 
break with sufficient force to prevent turtles from feeding on top. 

The growth of Caulerpa in the foraging habitat along Johnston's south 
shore is more difficult to characterize and define because, unlike 
Bryopsis , most of it cannot be seen from the surface. The extensive 
diving surveys with scuba devoted to examining habitat between the East 
and West Peninsulas helped to elucidate Caulerpa distribution. The macro- 
physa variety was far more abundant than uvif era or serrulata . The great- 
est Caulerpa coverage, approaching 100% on all hard substrate, occurred 
between the iron pilings and West Peninsula. Drifts of loose Caulerpa 
(again mostly macrophysa ) covered nearly all of the silt bottom areas 
between the pinnacles and other hard substrate. There was a considerable 
decline in Caulerpa growth seaward from the outer iron pilings and the 
end of West Peninsula. At a point <100 m from the end of West Peninsula, 
in the area of net location 2 (Fig. 1), the growth of Caulerpa nearly 
disappears. Throughout this transition zone, the Caulerpa coverage 
declines first on the seaward sides of the underwater pinnacles. With 
the disappearance of Caulerpa , the hard substrate consists of bare rock, 
an increasing amount of live corals, and the Bryopsis previously 
described. A consistent salinity of 34 c / oe was recorded from nine water 
samples taken throughout the area, including one taken at the greatest 
bottom depth of 8 m. The seawater temperature ranged from 27.5° to 
28.5°C. 

The findings in this study relating to Caulerpa coverage between the 
iron pilings and West Peninsula are considerably different from those 
presented by Applied Eco-Tech Services (1983). Figure 23 in their report 
shows 80% to 100% coverage of Caulerpa on hard bottom substrate extending 
200 m seaward of West Peninsula, twice the distance described in the 
present report. Possibly there wa6 a radical seasonal change, since the 
survey by Applied Eco-Tech Services (1983) was done in June 1983, 5 months 
before the present study. Confirmation of this important point is needed. 

The history of the marine environment off Johnston's south shore is 
virtually unknown due to the paucity of research conducted there. Sec- 
tions of marine habitat have been filled in with the island's periodic 
expansion. When Cris Balubar arrived in 1962, he recalls many turtles 
being present off the south shore, along with clear water and "nice 
reefs." He does not remember the status of algal cover. Lee Gohr helped 
to lay the existing sewer outfall in 1964 and recalls seeing benthic 
algae, but not in the density that exists at present. None of the person- 
nel interviewed was able to tell where the outfall existed before 1964. 

The exact role of sewage discharge in facilitating, or possibly 
depressing, algal growth remains speculative. Aerial photographs of the 
island taken over the years may help to answer this question. It may be 
that nutrients from guano in the brackish-water lens of the island his- 
torically served as fertilizers for algae. 



17 

Epizoites and Abnormalities 

Epizoic algal mats found mostly on the inguinal skin, plastron, and 
ventral surface of marginal scutes were sampled from five of the turtles 
captured. An analysis of thi6 material revealed eight kinds of algae, 
roundworms, foraminifera, amphipods, and black "mites" (Table 12). One 
of the algal species (Pilina sp.) may possibly be a new record for the 
tropical Pacific. The red alga, Polysiphonia tsudana, was present on all 
five turtles, and is also common on Hawaiian green turtles. 
Acrochaetium, Sphacelaria . and Lygbya have also been recorded on green 
turtles in Hawaii, but for the latter two genera as different species. 
Urospora sp., found on three turtles, has never been reported from 
turtles in Hawaii. Amphipods found on two turtles and in the mouth and 
esophagus of another turtle are probably Hyachelia tortugae (Tables 7 and 
8). This specialized crustacean, first described from the Galapagos 
Islands, has also been recorded in Hawaii and is known only from sea 
turtles. 

There was a noticeable paucity of barnacles on the turtles captured. 
None of the common turtle barnacle, Chelonibia testudinaria , was present, 
nor was the burrowing barnacle, Stephanolepas muricata . Even more sur- 
prising was the near absence of the harmless skin barnacle, Platylepas 
hexastylos . Only a few were found on two turtles and some of these were 
falling off and appeared to be dying. Platylepas commonly occurs in 
large numbers on the skin of Hawaiian green turtles. 

Small neoplastic growths 1-3 cm in diameter were found on two male 
turtles captured. Five growths were present on one and two on the other. 
The sites of these tumors included the trailing edge of front flippers, 
corner of the mouth, edge of the eye, inguinal region, and top of the 
head. Except for the latter location, growths are occasionally seen at 
these same places in Hawaiian and other green turtle populations. Some 
of the tumors documented elsewhere have been much larger, and far more 
numerous on individual turtles, than the ones seen at Johnston. However, 
Cris Balubar reported that about 10 years ago he did see a turtle 
"covered with white growths." The etiology of tumors on sea turtles is 
presently unknown. Relevant background material on the subject can be 
found in Harshbarger (1977), Balazs (1980c), Glazebrook et al. (1981), 
Jacobson (1981), and the references contained therein. 

Counts made of the major scales along the trailing edge of the front 
flippers, to a point perpendicular to the claw scale, revealed that 2 
(9.5%) of the 21 turtles varied from the standard 6 left-6 right count 
(Table 5). 

The number of postocular scales (scales posterior of the orbits) 
varied from the standard 4 left-4 right count in 4 (19%) of the turtles 
(Table 5). 

Four other abnormalities were noted among the turtles captured. The 
tail of an adult female was curled tightly under the edge of the carapace 
and appeared to be paralyzed. There was no scarring to suggest an 
injury . 



18 

One turtle made an unusual "whistling" sound when breathing. The cause 
could not be determined. Another had a small extra scute along the 
midline of the plastron between the humeral scutes, and another had a 
slightly depressed area to the carapace in the region of the 2d central 
scute and lst-2d right lateral scutes. There were no signs of an external 
injury. 

Predation and Injuries 

Evidence of injuries likely resulting from 6hark attack wa6 seen on 3 
of the 21 turtles captured. The injury on an adult male was the most 
pronounced: half of its tail was amputated and it had extensive, but mostly 
healed, lacerations of the right hind flipper. Two turtles had several 
deep cuts in their carapace. Some of these wounds, which were also mostly 
healed, could be seen near the painted numbers. Five other turtles had 
minor pieces missing from hind flippers, but this is common in green tur- 
tles. None of the turtles in Cris Balubar's photos showed signs of shark 
attack or other obvious injuries. The few probable shark-inflicted 
injuries seen at Johnston may not necessarily have happened there, since 
the adult turtles must periodically migrate elsewhere for breeding 
purposes. 

Sea turtles are known to be prey for large sharks, especially the 
tiger shark, Galeocerdo cuvier . Captain Cameron was well aware of thi6 
food preference when he used turtles as bait to enhance his shark fishing 
at Johnston in 1892 (Farrell 1928). The graphic description by Benson 
(1953) of a shark swallowing a turtle whole at Johnston also attests to 
this fact. More recently, C. B. Cecrle, in litt. 1979 cited by Balazs 
(1982d), reported sharks feeding on a large turtle outside of the atoll. 

Tiger sharks presently occur at the atoll, but they are not commonly 
seen nor frequently captured in recreational shark fishing. The most 
recent one caught was in August 1983 when a 4-m specimen was hooked in the 
ship channel at the western end of Sand Island. A 1-m whitetip reef shark, 
Triaenodon obseus , was used as bait. The stomach wa6 not cut open, so the 
natural prey items are unknown. 

The gray reef shark, Carcharhinus amblyrhynchos , often schools in 
abundance off the southwest corner of Johnston Island when kitchen garbage 
is dumped. Recreational fishermen fish for shark with baited hook6 from 
shore here and at West and East Peninsulas. However, no reports were 
received of turtle parts ever being found in the stomach of gray reef 
sharks. Elsewhere this species is apparently not a regular predator of 
turtles, except possibly on hatchlings. One person interviewed stated 
that he had seen two turtles feeding on dead fish used as bait on a shark 
hook at West Peninsula. 

Fishing for large tiger sharks used to be done with baited hooks set 
from markers along the main ship channel. This practice became prohibited, 
so personnel moored two large iron buoys 200-300 m offshore between West 
Peninsula and the southwest corner of the island. Both buoys are still 
present, but not often used for fishing since only small sharks (presumably 
gray reefs) have been caught there. During the diving surveys, only two 



19 

sharks were seen, a gray reef at location M, and a whitetip at location L 
(Fig. 2). Large numbers of the former are said to enter the atoll for 
breeding purposes during the summer months. 

The moray eel, Lycodontis javanicus (= Gymnothorax ). occurs at John- 
ston Atoll where it is known to be an opportunistic predator of fish, 
octopus, and spiny lobsters (Brock 1972). This species is the largest 
moray eel in the Indo-Pacif ic; specimens seen at Johnston are estimated at 
up to 2.4 m long (Randall 1980). Lycodontis javanicus could conceivably 
prey on small immature green turtles, but there is no direct evidence 
showing this. However, a prey item as large (and unusual) as a 46.5 cm 
whitetip shark was found by Brock (1972) in the stomach of a 1.4 m moray 
eel. Eels live under ledges and in other coral recesses similar to where 
green turtles are typically believed to sleep. The eel's abundance and 
prey items have not been intensively studied along the south shore of 
Johnston Island. None was seen during any of the diving surveys. How- 
ever, moray eels hide mostly in the reef during the daytime, and feed 
during twilight and at night. 

Large groupers (Serranidae) occur at Johnston, but there is presently 
no evidence that they prey on turtles. In Hawaii, and elsewhere, whole 
immature turtles have occasionally been found in the stomachs of these fish. 
Benson (1953) describes the capture of a stunned "430 lb (195 kg) black sea 
bass" following an explosion to clear coral in the lagoon. The contents of 
the stomach were not mentioned. Sometime during 1970-71, a 27 kg grouper, 
Epinephelus sp., was caught at the southwest corner of Johnston Island near 
the garbage chute. The stomach was full of spiny lobsters (R. E. Brock, 
pers. commun. , February 1984). 

An unusual relatively minor injury was found on four of the turtles 
captured. This consisted of an ulcer on the dorsal neck surface over the 
supraoccipital bone of the skull. The exact cause is unknown, but could 
involve repeated abrasion against hard substrate, perhaps the roof of a 
coral recess where the turtles sleep. 

Strandings 

Information was obtained on strandings of four turtles. Three were 
described in interviews with resident personnel, and one by correspondence 
from the U.S. National Museum concerning the specimen mentioned in Amerson 
and Shelton (1976). 

According to Cris Balubar, a large dead turtle "tangled in a Japanese 
fish net" washed up near East Peninsula several years ago. Also, at an 
unknown date a small turtle was found dead on the beach at Akau Island. No 
other information was available about these two incidents. 

Phil Roseberry related that a few years ago he was told by Army 
personnel about a large dead turtle which washed in at Akau Island. He 
visited the site the next day, but the turtle was gone, presumably washed 
away by the next high tide. 



20 

Dr. Jack G. Frazier provided the following data on file for the 
specimen at the U.S. National Museum: 

DSNM 163581 (Catalogue entry) 

Accession No. - 278016 

Original No. - 11287 

Name - Chelonia mydas/ japonica 

Locality - Sand Island, Johnston Atoll 

Date collected - November 20, 1966 

Received from - Smithsonian Institution Pacific Ocean 

Biological Survey Project 
Collected by - 

When entered - September 14, 1967 
Sex and No. of specimens - 1 head 

Dr. Frazier also related that the head is approximately 10 cm wide and 
preserved in alcohol. It appears to have been collected fresh, and has a 
cut on the left side similar to what would result from a blow with a 
machete. This is the extent of information available. A 10-cm wide head 
would have come from a turtle about 75 cm SCL, hence of an immature size. 
Cris Balubar was unaware of this specimen, as were the other longtime 
residents interviewed. A probable scenario might be that a freshly dead 
turtle washed up at Sand Island from the adjacent ship channel. A short 
time later it was found by Smithsonian personnel, who used a machete or ax 
to sever the head for the museum's collection. One of these cut6 may have 
accidently hit the side of the head. Other possible explanations might be 
that the gash resulted from a 6hark attack, or collision with a boat 
propeller. It seems unlikely that a live healthy turtle of this size would 
have been collected just for the head, unless the turtle was taken mainly 
for food or a trophy shell. By itself, and especially with no accompanying 
data, the head is of little taxonomic value, except to show the presence of 
Chelonia . 

The U.S. National Museum's subspecies designation of japonica is the 
name sometimes applied to western Pacific green turtles. The type locality 
is Japan (Hirth 1971). Based on existing information, the U6e of this or 
any other trinomial for the Johnston population is not currently justified. 

Basking 

No evidence was found during the field study that green turtles bask 
ashore during the daytime, as they commonly do at Johnston's closest island 
neighbor, French Frigate Shoals in the Northwestern Hawaiian Islands (Whit- 
tow and Balazs 1982). Balazs (1978) indicated that the turtle collected at 
Sand Island in 1966 may have been basking (or nesting). However, this now 
seems unlikely in view of the information presented in the previous section. 

Nitta (1982) stated that he received a report of turtles sometimes 
hauled out during the early morning on the coral-rubble and sand beach 
between the iron pilings and West Peninsula (Fig. 1). No turtles were 
found by Nitta (1982) when this site was visited at 0600 on July 9, 1982. 
Observers in the present study also surveyed the beach, which was 200 m 
long and accreted against the cement seawall that borders most of Johnston 



21 



Island. There were no marks in the beach suggestive of use by turtles. 
The site was visited on November 15 (2200), 16 (0500, and 2300) and 17 
(0500), but no turtles were seen. Cris Balubar reported that the beach is 
unstable in that portions often wash away and return. He sometimes visits 
here early in the morning to look for baitfish, but has never seen a turtle 
ashore. Nevertheless, it is conceivable that turtles do occasionally come 
out at night to sleep along this shoreline when underwater sleeping sites 
become unacceptable due to storm surge or other factors. Such behavior is 
known at Necker Island and Pearl and Hermes Reef (Northwestern Hawaiian 
Islands), and on the Na Pali coast of Kauai, Hawaii. 

An historic absence of basking by turtles at Johnston Atoll may have 
prompted the comment in 1859 by Brooke (MS.), quoted by Amerson and Shelton 
(1976), that "The reefs are covered with fish of various kinds. Mullet 
abound, but there are no turtles." Just 3 months earlier, in January 1859, 
Captain Brooke had visited French Frigate Shoals and Laysan Island, where 
he saw many turtles basking on the beach (Amerson 1971; Ely and Clapp 
1973). These sights might easily have caused him to believe that land 
basking was common behavior for all sea turtles at remote Pacific islands. 
Seeing none along the shoreline at Johnston during his 2-day visit would 
naturally have come as a surprise, hence causing him to comment about no 
turtles being there. Since turtles were caught at Johnston 33 years later 
in 1892 by Captain Cameron (Farrell 1928), Brooke's (MS.) statement could 
not have been applicable to turtles in the water. It most likely meant 
that he expected to see turtles on the beach, and that no special search 
was made for them in the waters of the atoll. It should also be noted that 
when Captain Brooke arrived, 14 men were already living on Johnston Island, 
some since July 1858, building a wharf and railway for shipping guano 
(Amerson and Shelton 1976). Captain Cameron noted the abandoned remains of 
this facility in 1892 (Farrell 1928). 

Reproduction 

No historic records are known of turtles nesting at Johnston Atoll, 
although Amerson and Shelton (1976) speculate that perhaps they did in 
small numbers before habitation by man. Johnston Island was inhabited 
intermittently for guano removal starting in July 1858. Sand Island, 
however, seems to have remained uninhabited until the mid-1930* s, when 
military construction began there. 

The nesting season for green turtles in northern latitudes is from May 
through August. At French Frigate Shoals (lat. 23°45'N), the major nesting 
site for Hawaiian green turtles, the peak months are June and July? A 
similar season should exist at Johnston (lat. 16°45'N), if turtles nested 
there. The Tanager Expedition spent 10 days in July 1923 camped on unin- 
habited Johnston Island. According to Amerson and Shelton (1976), Wetmore 
(MS. a and b) "...recorded no turtles." However, from this statement it is 
unclear if Alexander Wetmore, leader of the expedition, actually noted an 
absence of turtles, or just made no mention of them in his field journal. 
The latter case applies for Wetmore (MS. b), but I have not seen Wetmore 
(MS. a). Nevertheless, while camped on Johnston Island, Wetmore (MS. B) 
tells of a nightime walk along the beaches of "fine white coral sand." If 



22 

turtles were seen coming ashore to nest in numbers, surely it would have 
been mentioned. This should also have been the case for any extensive 
nesting excavations seen in the vegetation zone during numerous daytime 
surveys. However, it is conjecture if the same would also apply for Sand 
Island. There were no overnight camps there, and only two daytime visits 
were made. 

It should be noted that the Tanager Expedition had just come from 
French Frigate Shoals and other Northwestern Hawaiian Islands, where 
journal notes were in fact made about turtles basking, nesting, and being 
recently exploited (Amerson 1971). The expedition constituted the first 
discovery by scientific personnel of sea turtles actually basking. Conse- 
quently, most of the interest relating to turtles was focused on this 
behavioral aspect (Mellen 1925; Wetmore 1925). 

The question of whether or not turtles nested at Sand or Johnston 
Island when the Tanager Expedition visited in 1923 can probably be answered 
conclusively from aerial photographs. A small seaplane made the first 
photographic flights over the atoll during the expedition on July 12 and 19 
(Wetmore MS. b). These and other U.S. Navy aerial photos taken until the 
late 1930' s should be of sufficient resolution to show nesting excava- 
tions, if such sites existed. The original photos are apparently on file 
in the U.S. National Archives (Amerson and Shelton 1976). 

During his residency at Johnston Island since 1962, Cris Balubar only 
knows of one instance of a possible nesting, but couldn't recall the year. 
Turtle tracks and digging were seen at Sand Island, but there was no 
confirmation that eggs were laid. Cris Balubar also stated that he had 
never seen eggs inside turtles he had caught and cleaned. Upon further 
discussion, it was determined that only ova of a large and nearly mature 
size would have been noticed and remembered by him. 

Several people interviewed, including Cris Balubar, said that they 
had seen turtles locked together copulating, sometimes for extended 
periods. Cris Balubar felt that this mating activity was more prevalent 
during September through November. Francisco Aguinaldo and Ed Mattson saw 
a copulating pair in July 1983 at the shoreline of the sand beach by the 
iron pilings. The turtles swam off quickly when approached. Previous 
occurrences of this sort may have been partly responsible for the report 
received by Nitta (1982) of turtles hauling out on this beach. Present 
accounts of turtles copulating at Johnston are supported by the fact that, 
in 1892, Captain Cameron tethered a few receptive females to attract and 
catch males intent on mating (Farrell 1928). The contradiction here is 
that, for the most part, green turtles are thought to mate only in waters 
adjacent to their nesting beach. 

The lesions typically present on males and females following copula- 
tion were not seen on any of the turtles captured in this present study. 

A survey of Sand Island's shoreline identified four sand beaches, all 
located at the eastern end which comprised the original portion of the 
island before enlargement. Each of these natural beaches offers access to 
the elevated interior where the soil is suitable for nesting. With the 



23 

exception of one beach where there is a bright electrical light, Sand 
Island presently seems to be appropriate habitat for nesting, if turtles 
wanted to nest there. Even the beach by the iron pilings on Johnston 
Island has an elevated area of fine sand where nesting might successfully 
take place. In contrast, the man-made islands of Akau and Hikina are 
unsuitable, since the shorelines consist only of cement seawalls and coarse 
coral rubble, and the interiors are compacted aggregate from dredging. 

Migrations 

Since there presently is no nesting at Johnston Atoll, and possibly 
never was, the turtles must be periodically migrating elsewhere to repro- 
duce. This would be consistent with the pattern found in most other green 
turtle populations, where round trip migrations take place between resident 
foraging areas and distant breeding grounds. Migrations of this nature are 
known to be made, even when what may appear to be acceptable nesting 
beaches are close by. Though not conclusively proven, the turtles are most 
likely returning to breed at their natal beaches. 

A logical place where Johnston's turtles might go to nest would be 
French Frigate Shoals. This site is only 830 km to the north, and com- 
prises the major rookery for green turtles migrating from islands through- 
out a 2,200 km expanse of the Hawaiian Archipelago. However, there are no 
tag recoveries to support the hypothesis of French Frigate Shoals actually 
being the breeding site for Johnston's turtles. Ample opportunities have 
existed for such a migratory pattern to be demonstrated. Over the past 20 
years, more than 1,600 adults have been tagged at French Frigate Shoals, 
and about 200 others tagged elsewhere in the Northwestern Hawaiian Islands. 
None of these tagged turtles has been recovered in the Johnston population, 
neither in the present study, nor when turtle fishing was allowed at the 
atoll. Cris Balubar estimates he caught 60 turtles between 1967 and 1976, 
but found no tags. These same years were some of the most intensive for 
tagging turtles at French Frigate Shoals. Interestingly, during the years 
1966-72, 860 Hawaiian monk seals, Monachus schauinslandi , were also tagged 
in the Northwestern Hawaiian Islands. Three of these have subsequently 
been reported at Johnston; one came from French Frigate Shoals, and two 
from Laysan Island (Schreiber and Kridler 1969; Johnson and Kridler 1983). 
Monk seals were apparently a conspicuous component of Johnston's fauna when 
Captain Cameron visited there in 1892 (Farrell 1928). In this century, 
however, very few seals are known to have migrated outside the Hawaiian 
Archipelago, and then only to Johnston Atoll. A migratory link between 
Johnston and French Frigate Shoals also exists for seabirds. Of 733 inter- 
island tag recoveries, 32% involved French Frigate Shoals (Amerson and 
Shelton 1976). 

The isolated location of Johnston limits the number of other sites the 
resident turtles might go to nest. The Line Islands, starting with Kingman 
Reef 1,575 km to the southwest, is the next closest area to Johnston after 
French Frigate Shoals and the other Northwestern Hawaiian Islands. A low 
level of green turtle nesting occurs at the inhabited atolls of Fanning and 
Christmas in this island group. But none is known to occur at Kingman, nor 
the nearby atoll of Palmyra. 



24 

The closest island to the west of Johnston is uninhabited Bikar in 
the northern Marshall Islands. Bikar is located 2,220 km away and is 
considered a prominent nesting site for green turtles (Pritchard 1982). 
No tagging has been conducted at Bikar, so the resident foraging grounds 
for turtles nesting there are unknown. Presumably they include many of 
the Marshall Islands. Bikar is a traditional wildlife reserve owned by 
chiefs of a Marshallese clan who periodically visit to gather turtles, 
seabirds, and their eggs (Tobin 1952). 

Canton and other islands of the Phoenix group are situated almost 
directly south of Johnston, below the Equator, at a closest distance of 
2,160 km. Green turtles have been found to nest here, but again the 
areas where they migrate from are currently unknown (Balazs 1975). 

The limited amount of tagging done in the Pacific has uncovered some 
impressive long distance migrations by turtles, each encompassing the 
boundaries of several island nations (Balazs 1982c). Turtles tagged at 
Johnston in the present study may very well yield similar results, and 
eventually pinpoint the nesting site. 

From interviews, it was learned that three turtles had been marked at 
Johnston before the tagging done in this study. In 1971, Cris Balubar 
tagged a turtle (on the dorsal surface of a front flipper) with an aluminum 
plate embossed with his name, location (Johnston), weight of the turtle 
(130 lb (59.1 kg)), and date of release. The day and month could not be 
recalled. He also caught another turtle sometime in 1971 described as 
being "large and ugly," but apparently not diseased or injured. The cara- 
pace of this one was painted completely black before it was released. A 
second turtle was also marked with paint. This one weighed only about 14 
kg and was hand captured by someone else close to shore at Hikina Island. 
The words "Happy New Year January 1981" were stenciled on the carapace with 
fluorescent paint. No recoveries have been reported for these three 
turtles. 

Utilization 

It is not possible to quantify the number of turtles taken at Johnston 
over the years, except for the estimate of 60 given by Cris Balubar for 
1967-76. An estimate of 12-15 turtles harvested per year by "a long time 
resident" appears in Amerson and Shelton (1976). 

All turtles caught at the atoll since military occupation in the 
1930' s can be assumed to have been taken for sport, trophy shells, and as a 
seafood delicacy along with fish and lobster. There is no evidence of 
commercial harvest during this period, although some shells were probably 
sold privately. In 1976 the refuge manager for the U.S. Fish and Wildlife 
Service (Palmer Sekora) prohibited all taking of turtles at Johnston. This 
measure was instituted to achieve consistency with the already protected 
status of sea turtles at other National Wildlife Refuges in the United 
States. 

Cris Balubar 1 s case history of turtle fishing is worth describing, 
since he is well known locally for this skill. He appears to be the only 



25 

person over the past 22 years who regularly targeted turtles with a spe- 
cific fishing technique. This viewpoint could, however, be biased due to 
Cris Balubar 1 s long residency, and since he is still there to be inter- 
viewed. Other personnel fishing just for turtles may have come and gone 
over the years, leaving no verbal record of their activities. Turtle 
catching by others was probably only on an opportunistic basis while diving 
and spearf ishing. For instance, some years ago Lee Gohr caught a 59-kg 
turtle while recreational diving between Sand and North Islands. Also, 
since turtle fishing has been illegal since 1976, it is unlikely any infor- 
mation would be volunteered about turtles taken thereafter. A secondhand 
report, believed to be reliable, was received that occasionally turtles are 
still taken for food when they are accidentally encountered by divers in 
the north of the atoll. In these instances, it was reported the divers 
discard the shell and flippers before returning to Johnston Island. 

Cris Balubar' s turtle fishing started in 1967 when he was asked to 
catch a turtle for one of the island's military officers. As it is now, 
most turtles were found along the south shore. West Peninsula was gen- 
erally off limits for recreational purposes at that time. However, special 
permission for fishing was granted in this case. The technique success- 
fully employed involved a fiber glass pole with a spinner reel and treble 
hook to cast out and snag the turtle while it was feeding. A lengthy 
period would then often be needed to reel it in. The turtle incurred very 
little physical injury during capture, since it was usually hooked only in 
a flipper. This fishing method is a modification of one sometimes used for 
turtles in Hawaii (Balazs 1980c). Cris Balubar subsequently caught turtles 
in this manner from a small boat off the south shore, as well as from the 
seawall. He estimates that 45 of the 60 turtles taken were by boat. 
According to Balubar, Johnston turtles taste much better than Hawaiian 
turtles. This difference could be due to the algal food sources utilized. 
A recent article in Hawaii Fishing News describes his snagging technique 
used to land a large yellowfin tuna from the seawall at Johnston (Balubar 
1982). 

The only firm evidence of turtles being exploited at Johnston before 
military settlement is Captain Cameron's mention of their use in 1892 for 
shark bait and food (Farrell 1928). Turtles, along with seals, were a 
preferred bait for shark fishing at that time. It is reasonable to assume 
that other commercial fishing expeditions stopping there would also have 
taken turtles, if they could be found. In 1918 an attempt was made to 
establish a fishing station at Johnston, but dissatisfied workers quickly 
terminated the plan. During the 1920* s, at least two commercial fishing 
vessels visited the atoll. One was the Lanikai , a vessel known to take 
turtles from Pearl and Hermes Reef for markets in Honolulu (Thurston 1928; 
Amerson and Shelton 1976; Balazs 1980c). Brock et al. (1965) imply that at 
least one fishing boat from Honolulu regularly visited Johnston before 
World War II. 

Guano mining outposts on remote Pacific islands, like the one started 
at Johnston in 1858, regularly used whatever marine resources were avail- 
able for fresh food. Turtles were among those items eaten by guano diggers 
on Laysan Island. Presumably turtles were also eaten at other guano islands 
where they were present. Turtles were also eaten at the intermittent camps 



26 

set up illegally by plume hunters in the Northwestern Hawaiian Islands. 
The remains of such a camp were found at Johnston in 1923 (Wetmore MS. b). 

Environmental Perturbations 

A number of man-made perturbations have historically taken place at 
Johnston that could have adversely affected turtles. Some of these 
impacts, like blasting to remove coral, almost certainly caused direct 
mortality. Others have possibly been more subtle, thus making them diffi- 
cult to detect (Whittle et al. 1977). The highly modified environment at 
Johnston, and the various uncommon activities periodically conducted 
there, warrant the consideration of these perturbations. It is beyond the 
scope of this paper to give an in-depth evaluation and analyses of these 
factors. However, the information contained herein should provide some 
direction for possible future research of turtles at Johnston Atoll. 

Initial blasting and dredging to clear coral and increase the size of 
Johnston and Sand Islands happened from 1939 to 1942. Woodbury (1946) 
describes the difficulties experienced in breaking up the coral, and the 
extraordinary amount of dynamiting required. Johnston Island was expanded 
from its original 19 ha to 85 ha, and 4-ha Sand Island was doubled in size. 
Major construction projects to further enlarge Johnston Island were com- 
pleted in 1958 and 1964, resulting in the present land area of 253 ha. The 
man-made islands of Akau (10 ha) and Hikina (7 ha), with channels leading 
to them, were also finished in 1964 (U.S. Army Corps of Engineers 1983). 
The concussions from underwater explosions during these various projects 
would have 6tunned any turtle in the vicinity causing injury and death from 
direct impact or drowning. Such blasting, and the resulting dead fish, 
would also be an attractant to large sharks that could, in turn, easily 
have preyed on stunned turtles. The likelihood that this actually happened 
is supported by Benson's (1953) description of "...numerous huge sharks" 
being present during dynamiting, and one of them seen "...swallowing a sea 
turtle whole." Underwater explosions have also been set off occasionally 
in recent years. Moses Caballero mentioned the destruction of a live bomb 
found in the turning basin of the ship channel. It should be noted that 
some blasting also occurred in the lagoon waters of Johnston long before 
military activities. During Captain Cameron's 1892 visit, at least one 
marine harvest was made by driving fish close to shore and setting off 
"dynamite cartridges" (Farrell 1928: p. 406). 

Several hydrogen bombs were successfully detonated at high altitude 
over Johnston in 1958 and 1962. Amerson and Shelton (1976) stated the 
following with respect to one of the environmental impacts: 

"During these nuclear tests, an elaborate water sprinkler system was 
installed on the original portion of Sand Island to protect the birds 
living there. In addition, other protective devices were used, 
including smoke pots placed upwind as a shade screen and aerial flares 
to divert the birds' attention from the flash of the blast itself." 

Since the flash (and heat?) from these explosions was so intense at 
ground level, eye damage could presumably have resulted to any turtles on 
or near the surface at the time of detonation. 



27 

During 1962, three of the missiles carrying nuclear devices malfunc- 
tioned, exploded, and spread particles of radioactive material over the 
atoll and surrounding Pacific Ocean. One of these blasts was on the launch 
pad, significantly affecting a portion of Johnston Island (Anonymous 1962a, 
1962b, 1962c). Decontamination efforts have been periodically carried out, 
including removal and dumping of debris and surface soil over deep water 
outside the atoll. As a result, some contaminated soil entered the near- 
shore waters of the island and certain areas of the island continue to be 
"off limits." The surface runoff of rainfall undoubtedly transports addi- 
tional particles into the nearshore waters. The soil at the western end of 
Sand Island is also still contaminated (U.S. Army Corps of Engineers 1983). 
Benthic algae, including types which exist in the turtle foraging habitat 
along the south shore, concentrate certain radionuclides at rates higher 
than other plants and animals (Hines 1962: p. 145-151; Whittle et al. 
1977). Hillestad et al. (1974) reported very low levels of gamma emitters 
in tissue from loggerhead turtles, Caretta caretta , in Georgia and South 
Carolina. The levels present in turtles from a contaminated area are 
apparently unknown. 

During the early 1970* s, various chemical weapons and herbicides were 
stored on Johnston Island. Also, canisters of dioxin were flown to the 
atoll in 1976 (Benson 1976). The herbicide was safely incinerated at sea 
in 1977 using a ship built for such purposes (Nelson 1977). However, 
leakage from these drums resulted in Herbicide Orange and dioxin contami- 
nation of soil at the northwest corner of the island. In 1973, various 
water, sediment, and marine biological samples were tested for dioxin and 
Herbicide Orange. A muscle and liver specimen from a green turtle of 
unknown size, taken between East and West Peninsulas, were included. A 
sample of Bryopsis , but not Caulerpa , from the same area was also anal- 
yzed. None of these four samples from 1973 (presumably measured on a dry 
matter basis) was found to have detectable levels of dioxin or Herbicide 
Orange (U.S. Army Corps of Engineers 1983: footnote in Table III-6, 
Appendix L). 

Some leakage of nerve gas from deteriorating munitions containers has 
been reported (Borg 1982). As noted in a previous section, wash water 
from the decontamination procedure is discharged through the sewer outfall 
directly into the turtle foraging habitat. Storm drains may also receive 
some of this effluent. 

A considerable amount of man-made debris was seen on the ocean bottom 
during diving surveys at locations G and U (Fig. 2). These were the only 
two offshore (>14 m) sites surveyed, so discarded items may exist over a 
much broader area off Johnston Island's south shore. The material seen 
consisted of 55-gal drums, a trailer, a Mike boat, and pieces of iron 
reinforcement bar. The drums were heavily rusted, but some appeared 
intact. The contents, if any, could not be determined through interviews 
with resident personnel. 

Heavy metals are known to be discharged from desalination plants as 
the result of internal corrosion. Two types of discharge have been 
reported: one emitted when the plant is operating normally, the other 
produced during periodic cleaning and maintenance cycles (Chesher 1975). 



28 

High levels of copper are present in the normal effluent, and this element 
is believed to be the most toxic to marine organisms. When desalination 
plants shut down for maintenance, corroded copper-nickel surfaces dry and 
oxidize. With resumption of operation, copper contamination is 2-3 times 
higher than normal for a few hours. Higher levels of nickel and iron are 
also released. The discharge during this period is turbid and black 
(Chesher 1975). Few studies have attempted to measure heavy metal content 
in sea turtles and their eggs (Hillestad et al. 1974; Stoneburner et al. 
1980; Witkowski and Frazier 1982). Furthermore, as emphasized by Witkow- 
ski and Frazier (1982) and Coston-Clements and Hoss (1983), it is diffi- 
cult to determine the significance of such findings because little is 
known about baseline levels and physiological effects. Based on reports 
of copper levels discharged from a desalination plant (Chesher 1975), the 
facility at Johnston probably produces at present 1.3-2.6 kg of copper 
effluent per day under normal operation. Effluent from the plant appar- 
ently also serves to enrich, by some undetermined process, the existing 
radionuclide contamination in nearshore waters, thereby producing a local- 
ized "hot spot" (U.S. Army Corps of Engineers 1983: Appendix L, p. 5-6). 
Whether or not cooling water from other facilities besides desalination 
would do the same is unknown. 

Petroleum spills can adversely affect turtles by external fouling, 
ingestion, and interference with olfactory perception and food supply 
(Coston-Clements and Hoss 1983). During the field study at Johnston, 
dried petroleum matter was found adhering to the seawall at the east 
corner of West Peninsula. It had likely gathered and washed up there from 
the funneling effect of prevailing winds and currents. The age of the 
material, and the length of time required for it to accumulate, could not 
be determined. Turtle foraging habitat around West Peninsula, and along 
much of the south shore, appears to be vulnerable to petroleum contamina- 
tion due to its windage and proximity to the ship channel. 

Artificial illumination on beaches is known to discourage adult 
turtles from nesting and disorient hatchlings crawling to the sea (Coston- 
Clements and Hoss 1983). However, almost no information exists on the 
effects of coastal lights on turtles foraging or sleeping at night in 
nearshore habitat. Nocturnal feeding is common behavior for green turtles 
in Hawaii (Balazs 1980c). However, at Johnston the catch rates from nets, 
and direct observations made from shore, suggest that foraging is mostly, 
if not entirely, during the daytime. Between West Peninsula and the 
island's southwest corner there are nine white lights of medium intensity 
set on posts 75-100 m inland. In addition, 23 dim yellow lights are 
located on the chemical storage bunkers facing the shoreline. None of 
these lights directly illuminate the nearshore waters, although they are 
clearly visible from offshore. At present, there are no lights on posts 
anywhere near the shoreline of West Peninsula itself. 

Cholinesterase 

Very low concentrations of organophosphorus compounds inhibit the 
activity of cholinesterase, an enzyme responsible for important physio- 
logical processes in the nervous system. Cholinesterase also occurs in 
serum and red blood cells. The inhibitory effect of organophosphorus 



29 

compounds is the basis for their use as insecticides and certain chemical 
weapons. Cholinesterase inhibition can also be used to biochemically 
detect organophosphorus compounds in the environment or in an organism 
(Namba 1971; Lundin 1975). 

At Johnston Island, red blood cell cholinesterase is measured in 
humans by the 17-Minute Manual Method. According to information supplied 
by Lucille Bodnar, who routinely performs this analysis at Johnston, the 
method is based on the principle that cholinesterase hydrolyzes acetyl- 
choline bromide with the production of acetic acid. The change in pH is 
measured in a barbital-phosphate buffer when red blood cells are mixed 
with a known excess of acetylcholine bromide and allowed to hemolyze. The 
results are expressed in terms of the decrease in pH units during the 17- 
minute reaction period. The normal range for humans is 0.63-0.89 pH/h. 

Cholinesterase values, using a 0.2 ml aliquot of red blood cells, 
were determined by this method for six adult and three immature turtles 
(Table 13). The validity of these results is unknown, since the analysis 
is designed for human blood. The method may be unsuitable for turtle 
blood due to the capacity of the buffer, size of aliquot used, or other 
factors. Verification is needed. In addition, the normal range of 
decrease in pH units for the green turtle is currently unknown. Given 
these analytical uncertainties and the small numbers sampled, Table 13 
shows that cholinesterase for the nine turtles ranged from 0.13 to 0.34 
pH/h. It is worth noting that the mean value for the three adult males 
(0.27 pH/h) was almost double the mean of the three adult females (0.14 
pH/h). Also, the two i mma ture females measured had levels similar to 
the adult males. Two of the nine turtles sampled were caught at net 
location 7, downwind of the sewer outfall. There is no indication from 
these few data that differences exist between sampling locations. 

RECOMMENDATIONS 

Management Measures 

The information contained in this paper provides a basis for offering 
recommendations of management measures to help ensure the conservation of 
turtles at Johnston Atoll. These actions are: 

1. A specific management zone for marine turtles should be estab- 
lished. The area should encompass marine habitat extending sea- 
ward for about 1 km along the entire south shore of Johnston 
Island, as well as a contiguous band extending about 1.5 km to the 
northeast of the main ship channel. The purpose of this zone 
would be to give special attention to the turtles concentrated 
there and the habitat upon which they depend. An appropriate and 
distinct mechanism would then exist to soundly manage the area on 
a continuing basis. The designation would be particularly helpful 
for identifying and evaluating any potential impacts to turtles 
and habitat that might arise in the future. The zone would be 
fully consistent with the environmental goals of the JACADS pro- 
ject and, in fact, the project would likely benefit from the 
special management attention given to the turtles. 



30 



2. A management action needed at present is the curtailment of any 
recreational boats transiting or anchoring in the area described 
above. The rapid diving response when turtles are approached by 
boats indicates that normal foraging behavior is easily disrupted. 
This may be the result of previous human harassment, including 
fishing efforts to hook them and regular encounters with small 
boats. 

3. A formal system should be implemented to deal with any future 
strandings of dead or live turtles. Rapid reporting, and the 
appropriate immediate response by interested parties, is 
absolutely essential for these cases. Valuable specimens and 
data can be acquired in this manner; for example, bones for age 
determination, whole stomach contents, tissue samples, and a 
determination of the cause of death or debilitation. The 
presence of a tag further increases the worth of the specimen. 
The system should also include turtles or their parts found in 
the stomach of sharks and other predators. 

4. An informative, interesting, and inexpensive brochure, preferably 
with illustrative photographs, should be prepared telling about 
the turtles at Johnston, where they principally occur, and their 
protected status under the U.S. Endangered Species Act. The 
brochure should be specific for turtles, and not done in 
descriptive combination with other wildlife or marine resources 
of the atoll. The brochure should be distributed at the air 
terminal to each new person upon arrival. 

5. A formal response plan should be prepared describing the actions 
to be taken in the event of a petroleum spill involving the area 
described for a turtle management zone. Special attention should 
be given to sites around West Peninsula where spillage may 
concentrate. 

6. A plan to assess the effects, if any, of newly installed lights 
on the foraging behavior and other use patterns of green turtles 
off West Peninsula should be developed. This should encompass 
the temporary lights needed during active construction of JACADS, 
as well as permanent security lights planned for the completed 
facility. 

Future Research Activities 

The successful long-term management of these turtles is, to a large 
extent, dependent upon a certain amount of future research being 
accomplished. Research on turtles at Johnston has long been neglected. 
However, from this present assessment it is apparent that they constitute 
an ecologically important, scientifically challenging, and historically 
interesting part of the atoll's fauna. In addition, Johnston's turtles 
are most likely used for food by native people somewhere in the Pacific 
islands, since it is doubtful they nest at French Frigate Shoals where 
full protection would be afforded. A major research and management goal 



31 

should be to determine the international migrations of these turtles, 
including their ultimate destination and island areas of transit where 
fishing may occur. The only way to achieve this objective at an early 
date is to capture and tag more turtles at Johnston. The relatively high 
proportion of adults and females found in the population will be an advan- 
tage to understanding the movement patterns, since it will increase the 
probability of long-distance recoveries. 

The following recommendations relate to research that should be 
accomplished to facilitate a better understanding of the biology of this 
turtle population. The information developed in these studies will also 
serve as a basis to formulate future management measures for Johnston's 
turtles. While this research is clearly needed, it is outside the scope 
of this paper to indicate specific agency responsibility or priorities 
for support of this work. 

1. A standard monitoring program should be established to assess and 
tag turtles periodically in a manner similar to the present 
study. This action will be particularly important during the 
active construction phase of the JACADS project. During this 
period, three 10-day study visits per year are deemed necessary. 
Thereafter, one or two visits per year would be sufficient. 

2. Diving surveys with scuba should be made between West Peninsula 
and the southwest corner of Johnston Island to search for turtle 
sleeping areas. To accomplish the dives safely, formal arrange- 
ments must be made to delay, for 2 h daily, the interval pumping 
of sewage from the outfall over a 3-4 day period. This appears 
feasible at present during midmorning when water usage is nor- 
mally low. However, it must be done before the large increase in 
personnel scheduled for the JACADS project. 

3. The blood analysis used in the present study to measure 
cholinesterase should be evaluated and, if needed, modified to 
obtain accurate measurements. Routine testing of cholinesterase 
in turtles should be conducted as part of the periodic monitoring 
suggested in recommendation 1 above. The normal range for green 
turtles should be determined from blood sampling currently under- 
way in Hawaii. 

4. The enrichment of radionuclide contamination by effluent from the 
desalination plant should be elucidated. The possible role of 
heat and heavy metals in this process should be examined to 
ascertain if discharge water planned for JACADS will produce 
similar enrichment, which in turn may be transferred to turtles 
through algal food sources. 

5. Aerial photographs taken over Johnston Atoll should be located 
and examined to determine the past distribution of benthic algae 
and if nesting occurred during the period before large scale 
inhabitation by man. 



32 

ACKNOWLEDGMENTS 

A number of individuals contributed substantially to the success of 
this research project. Algae samples were identified by Dennis J. 
Russell, Department of Biology, Seattle Pacific University, Seattle, 
Washington, who also supplied information on the habitat requirements and 
other ecological considerations of benthic algae occurring at Johnston 
Atoll. The nutrient and mineral composition of Bryopsis and Caulerpa was 
determined by Stanley Ishizaki of the Feed and Forage Analyses Program, 
University of Hawaii, Honolulu, Hawaii. Testosterone levels were deter- 
mined by Thane Wibbels and David W. Owens, Department of Biology, Texas 
A&M University, College Station, Texas. Cholinesterase activity was mea- 
sured by Lucille Bodnar of the Johnston Island medical facility. Data on 
the turtle specimen stored at the U.S. National Museum was supplied by 
Jack G. Frazier. 

Grateful appreciation is also expressed to the numerous persons who 
provided valuable and previously unrecorded information about sea turtles 
and related aspects of Johnston Atoll. As evident from the contents of 
this paper, Cris Balubar was particularly helpful in contributing his 
local knowledge and past experiences. 

Lt. Col. Patrick C. Moore (Commanding Officer), Lt. Col. R. H. Jolley, 
Major J. T. Mitchell, Major J. D. Aiken, CWO Gary G. Stillman and other 
members of the military services at Johnston are to be sincerely thanked 
for their assistance, advice, and hospitality during the course of the 
field work. Gratitude is also extended to John M. Merle, Resident Mana- 
ger, and other personnel of Holmes and Narver Inc. , for their fine assis- 
tance and cooperation. 

James Maragos and Bill Lennan of the U.S. Army Corp6 of Engineers, 
Pacific Ocean Division, provided considerable coordination, help and gui- 
dance during all phases of the project. 

The field work at Johnston was allowed under an Area Clearance from 
the Defense Nuclear Agency, and by a Special Use Permit (JHN-2-83) issued 
by the U.S. Fish and Wildlife Service. 



LITERATURE CITED 

Amerson, A. B. , Jr. 1971. The natural history of French Frigate Shoals, 
Northwestern Hawaiian Islands. Atoll Res. Bull. 150:1-383. 

Amerson, A. B., Jr., and P. C. Shelton. 1976. The natural history of 

Johnston Atoll, central Pacific Ocean. Atoll Res. Bull. 192:1-479. 

Anonymous. 1962a. Warhead detonation is averted. The Honolulu 
Advertiser, July 27, 1962. 

1962b. Site repairs may take up to eight weeks. Honolulu 
Star-Bulletin, July 31, 1962, p. 1. 



33 



Anonymous. 1962c. Destroys Johnston N-shot. The Honolulu Advertiser, 
Oct. 16, 1962, p. 1-2. 

Applied Eco-Tech Services, Inc. 1983. Johnston Atoll water quality and 
current study - a final report. Report submitted to U.S. Army 
Engineer District, Honolulu, 24 Aug. 1983, 100 p. 

Balazs, G. H. 1975. Marine turtles in the Phoenix Islands. Atoll Res. 
Bull. 184:1-7. 

1976. Green turtle migrations in the Hawaiian Archipelago. Biol. 
Conserv. 9:125-140. 

1978. Terrestrial critical habitat for sea turtles under United 
States jurisdiction in the Pacific region. 'Elepaio 39(4):37-41. 

1980a. Field methods for sampling the dietary components of green 
turtles, Chelonia mvdas . Herpetol. Rev. 11(1) :5-6. 

1980b. A review of basic biological data on the green turtle in the 
Northwestern Hawaiian Islands. In R. W. Grigg and R. T. Pfund (edi- 
tors), Proceedings of the Symposium on Status of Resource Investiga- 
tions in the Northwestern Hawaiian Islands, April 24-25, 1980, Univer- 
sity of Hawaii, Honolulu, Hawaii, p. 42-54. Sea Grant Misc. Rep. 
UNIHI-SEAGRANT-MR-80-4 . 

1980c. Synopsis of biological data on the green turtle in the 
Hawaiian Islands. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, NOAA-TM- 
NMFS-SWFC-7, 141 p. 

1982a. Factors affecting the retention of metal tags on sea turtles. 
Mar. Turtle Newsl. 20:11-14. 

1982b. Growth rates of immature green turtles in the Hawaiian Archi- 
pelago. In K. A. Bjorndal (editor), Biology and conservation of sea 
turtles, p. 117-125. Smithson. Inst. Press, Wash., D.C. 

1982c. Sea turtles: A shared resource of the Pacific islands. South 
Pac. Coram. Fish. Newsl. 23:22-24. 

1982d. Status of sea turtles in the central Pacific Ocean. In K. A. 
Bjorndal (editor), Biology and conservation of sea turtles, p. 243- 
252. Smithson. Inst. Press, Wash., D.C. 

1983. Recovery records of adult green turtles observed or originally 

tagged at French Frigate Shoals, Northwestern Hawaiian Islands. 

U.S. Dep. Commer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SWFC-36, 42 p. 

Balubar, C. 1982. Biggest fish story — mail buoy. Hawaii Fish. News 
7(10):8. 



34 

Benson, B. 1953. Johnston Island — it's not exactly paradise. Honolulu 
Star-Bulletin, Nov. 8, 1953, magazine section, p. 15. 

1976. Air Force flies deadly dioxin canisters to Johnston Atoll. The 
Honolulu Advertiser, Dec. 9, 1976, A: 14. 

Bentley, T. B. , and A. Dunbar-Cooper. 1980. A blood sampling technique 
for 6ea turtles. Final report for contract No. NA-80-GE-A-00082, 
Southeast Fisheries Center, National Marine Fisheries Service, NOAA, 
14 p. 

Borg, J. 1982. Nerve gas leaks seen under control. The Honolulu 
Advertiser, Nov. 4, 1982, A:5. 

Brock, R. E. 1972. A contribution to the biology of Gymnothorax iavanicus 
(Bleeker). M.S. Thesis, Univ. Hawaii, 121 p. 

Brock, V. E. , R. S. Jones, and P. Helfrich. 1965. An ecological recon- 
naissance of Johnston Island and the effects of dredging. Univ. 
Hawaii, Hawaii Mar. Lab., Tech. Rep. 5, 90 p. 

Brock, V. E. , W. Van Henkelem, and P. Helfrich. 1966. An ecological 

reconnaissance of Johnston Island and the effects of dredging. Second 
Annual Report. Univ. Hawaii, Hawaii Mar. Lab., Tech. Rep. 11, 56 p. 

Brooke, J. M. MS. Journal of the Fenimore Cooper for 1859. [Unpublished 
manuscr. ] G. M. Brooke, Jr., Lexington, Va. 

Buggeln, R. G. , and R. T. Tsuda. 1966. A preliminary marine algal flora 
from selected habitats on Johnston Atoll. Univ. Hawaii, Hawaii Inst. 
Mar. Biol., Tech. Rep. 9, 29 p. 

1969. A record of benthic marine algae from Johnston Atoll. Atoll 
Res. Bull. 120:1-20. 

Carr, A. 1972. Great reptiles, great enigmas. Audubon 74(2): 24-35. 

Chesher, R. H. 1975. Biological impact of a large-scale desalination 

plant at Key West, Florida. In E. J, Ferguson Wood and R. E. Johannes 
(editors), Tropical marine pollution, p. 99-153. Elsevier Oceanogr. 
Ser. 12. 

Coston-Clements, L. , and D. E. Hoss. 1983. Synopsis of data on the impact 
of habitat alteration on sea turtles around the southeastern United 
States. U.S. Dep. Commer., NOAA Tech. Memo. NMFS, N0AA-TM-SEFC-117, 57 p. 

Dawes, C. J., and R. H. Goddard. 1978. Chemical composition of the wound 
plug and entire plants for species of the coenocytic green alga, 
Caulerpa . J. Exp. Mar. Biol. Ecol. 35:259-263. 

Ely, C. A., and R. B. Clapp. 1973. The natural history of Laysan Island, 
Northwestern Hawaiian Islands. Atoll Res. Bull. 171:1-361. 



35 

Farrell, A. (transcriber). 1928. John Cameron's odyssey. The MacMillan 
Co., 461 p. 

Glazebrook, J. S. , R. S. F. Campbell, and D. Blair. 1981. Pathological 
changes associated with cardiovascular trematodes (Digenea: 
Spirorchidae) in a green sea turtle Chelonia my das (L). J. Comp. 
Pathol. 91:361-368. 

Harshbarger, J. C. 1977. Role of the registry of tumors in lower animals 
in the study of environmental carcinogenesis in aquatic animals. In 
H. F. Kraybill, C. J. Dawe, J. C. Harshbarger, and R. J. Tardiff 
(editors), Aquatic pollutants and biologic effects with emphasis on 
neoplasia, p. 280-289. New York Acad. Sci., N.Y. 

Hillestad, H. 0., R. J. Reimold, R. R. Stickney, H. L. Windom, and J. H. 
Jenkins. 1974. Pesticides, heavy metals and radionuclide uptake in 
loggerhead sea turtles from South Carolina and Georgia. Herpetol. 
Rev. 5(3):75. 

Hines, N. 0. 1962. Proving ground. Univ. Washington Press, Seattle, 366 p. 

Hirth, H. F. 1971. Synopsis of biological data on the green turtle 
Chelonia mydas (Linnaeus) 1758). FAO Fish. Synop. 85, 1:1-8:19. 

Ireland, L. C. 1979a. Homing behavior of immature green turtles ( Chelonia 
mydas ) . Am. Zool. 19:952. 

1979b. Homing behavior of juvenile green turtles, Chelonia mydas . In 
C. J. Amlaner, Jr., and D. W. MacDonald (editors), A handbook of 
biotelemetry and radio tracking, p. 761-764. Pergamon Press, Oxford. 

Jacobson, E. R. 1981. Virus associated neoplasms of reptiles. In C. J. 
Dawe, J. C. Harshbarger, S. Rondo, T. Sugimura, and S. Takayama (edi- 
tors), Phyletic approaches to cancer, p. 53-58. Jpn. Sci. Soc. Press, 
Tokyo. 

Johnson, A. M. , and E. Kridler. 1983. Interisland movement of Hawaiian 
monk seals. 'Elepaio 44(5): 43-46. 

Ludwig, G. M. 1982. Trip report: Johnston Atoll, September 12-17, 1981. 
U.S. Fish and Wildlife Service, Honolulu, 20 p. 

Ludwig, G. M. , S. Fefer, J. Maragos, and E. Nitta. 1982. JACADS project 
environmental survey at Johnston Atoll NWR July 8-11, 1982. Trip 
report. U.S. Fish Wildl. Serv. , Honolulu, 21 p. 

Lundin, S. J. 1975. The inhibition of cholinesterase activity by organo- 
phosphorus compounds as a means of an inspection procedure. In The 
problem of chemical and biological warfare. Vol. VI. Technical 
aspects of early warning and verification, p. 177-183. Humanities 
Press, N.Y. 



36 

Meinesz, A. J. Jaubert, and M. Denizot. 1981. Distribution of the algae 

belonging to the genus Caulerpa in French Polynesia (atoll of Takapoto 
and island of Moorea). Proceedings of the Fourth International Coral 
Reef Symposium, Manila 2:431-437. 

Mellen, I. M. 1925. Marine turtles sleep on Hawaiian sands. N.Y. Zool. 
Soc, Bull. 28:160-161. 

Menzies, R. A., L. Kochinsky, J. M. Kerrigan. 1983. Techniques for muscle 
biopsy and blood sampling from sea turtles. Poster session presented 
at the Western Atlantic Turtle Symposium, July 17-22, 1983, San Jose, 
Costa Rica, 2 p. 

Namba, T. 1971. Choi inest erase inhibition by organophosphorus compounds 
and its clinical effects. Bull. WHO 44:289-307. 

Nelson, L. 1977. Herbicide burning to end this week. Honolulu Star- 
Bulletin, July 27, 1977, C:4. 

Nitta, E. T. 1982. Johnston Atoll field trip report, July 1982. South- 
west Region, Western Pacific Program Office, Natl. Mar. Fish. Serv., 
NOAA, Honolulu, 6 p. 

Owens, D. W. , and G. J. Ruiz. 1980. New methods of obtaining blood and 
cerebrospiral fluid (from marine turtles. Herpetologica 36:17-20. 

Paul, V. J. 1983. Strategies of chemical defense in tropical marine 
algae of the order Caulerpales (Chlorophyta). (Abstract.) 64th 
Annual Meeting of the Western Society of Naturalists, Burnaby, British 
Columbia. 

Paul, V. J., and W. Fenical. 1982. Toxic feeding deterrents from the 

tropical marine alga Caulerpa bikinensis (Chlorophyta). Tetrahedron 
Lett. 23, 48:5017-5020. 

Pritchard, P. C. H. 1982. Marine turtles of the South Pacific. In K. A. 
Bjorndal (editor), Biology and conservation of sea turtles, p. 253- 
262. Smithson. Inst. Press, Wash., D.C. 

Rainey, W. E. 1981. Guide to sea turtle visceral anatomy. U.S. Dep. 
Commer., NOAA Tech. Memo. NMFS, NOAA-TM-NMFS-SEFC-82, 82 p. 

Randall, J. E. 1980. A survey of ciguatera at Enewetak and Bikini, Mar- 
shall Islands, with notes on the systematics and food habits of cigua- 
toxic fishes. Fish. Bull., U.S. 78:201-249. 

Russell, D. J., and B. A. Carlson. 1978. Edible-oil pollution on Fanning 
Island. Pac. Sci. 32:1-15. 

Schreiber, R. W. , and E. Kridler. 1969. Occurrence of an Hawaiian monk 
seal ( Monachus schauinslandi ) on Johnston Atoll, Pacific Ocean. J. 
Mammal. 50:841-842. 



37 

Stoneburner, D. L. , M. N. Nicora, and E. R. Blood. 1980. Heavy metals in 
loggerhead sea turtle eggs ( Caretta caretta ). Evidence to support the 
hypothesis that demes exist in the western Atlantic population. J. 
Herpetol. 14:171-175. 

Thurston, L. A. 1928. Features of Johnston Island. The Honolulu 
Advertiser, Aug. 5, 1928, p. 4, 11. 

Tobin, J. E. 1952. Land tenure in the Marshall Islands. Atoll Res. Bull. 
11:1-36. 

U.S. Army Corps of Engineers. 1983. Johnston Atoll chemical agent dis- 
posal system (JACADS) - Final Environmental Impact Statement, 1 Nov. 
1983. Pacific Ocean Division, Corps of Engineers, Ft. Shafter, 
Hawaii, 75 p. + Appendices A-M. 

U.S. Fish and Wildlife Service. MS. Pacific Islands National Wildlife 
Refuges: Johnston Atoll NWR, Baker Island NWR, Howland Island NWR, 
Jarvis Island NWR. U.S. Fish and Wildlife Service, Honolulu, Hawaii. 

Votaw, H. C. 1943. Johnston Island. U.S. Naval Institute Proceedings, 
69, 487:1176-1180. 

Wetmore, A. 1925. Bird life among lava rock and coral sand. Natl. Geogr. 
Mag. 59:76-108. 

MS. a. Field notes, 1923. [Unpublished manuscr. ] Smithson. Inst., 
Wash. D.C. 

MS. b. A scientific survey of Johnston Island, 1923, 25 p. [Unpub- 
lished manuscr. prepared in 1963.] [Page 5 missing from copy seen on 
file at the U.S. Fish and Wildlife Service, Honolulu.] 

Whittle, K. J., R. Hardy, A. V. Holden, R. Johnson, and R. J. Pentreath. 
1977. Occurrence and fate of organic and inorganic contaminants in 
marine animals. In H. F. Kraybill, C. J. Dawe, J. C. Harshbarger, and 
R. J. Tardiff (editors), Aquatic pollutants and biologic effects with 
emphasis on neoplasia, p. 47-79. New York Acad. Sci., N.Y. 

Whittow, G. C, and G. H. Balazs. 1982. Basking behavior of the Hawaiian 
green turtle ( Che Ionia mydas ). Pac. Sci. 36:129-139. 

Witkowski, S. A., and J. G. Frazier. 1982. Heavy metals in sea turtles. 
Mar. Pollut. Bull. 13:254-255. 

Woodbury, D. 0. 1946. Builders for battle. E. P. Dutton and Co., N.Y. , 
415 p. 



38 



Table 1. --Results of turtle netting effort in meter-hours (MH). 







Phase 


1 




Phase 


2 




Total 




Net 
loca- 






















MH per 


Number 




MH per 


Number 




MH per 


Number 


tion 


MH 


turtle 


captured 


MH 


turtle 


captured 


MH 


turtle 


captured 


1 


4,460 


1,115 


4 


1,975 


1,975 


1 


6,435 


1,287 


5 


2 


509 







4,206 


526 


8 


4,715 


589 


8 


3 


1,085 


362 


3 


3,513 







4,598 


1,533 


3 


4 


567 







— 




— 


567 







5 


1,026 







— 




— 


1,026 







6 


216 







— 




— 


216 







7 


— 




— 


2,748 


550 


5 


2,748 


550 


5 


8 


— 




— 


280 







280 







9 


— 




— 


400 







400 







10 


— 




— 


340 







340 







11 


— 




— 


949 







949 







12 


— 




— 


820 







820 







13 


— 




— 


153 







153 







14 


— 




— 


340 







340 







15 


— 




— 


420 







420 







16 


— 




— 


420 







420 







17 


— 




— 


189 







189 







Total 


7,863 


1,123 


7 


16,753 


1,197 


14 


24,616 


1,172 


21 



Table 2. — Daily turtle netting effort. 



Date 
1983 



Net 
location 



Number 
captured 



Duration 
in hours 



Length of 
net (■) 



Netting effort 
(meter-hour b ) 



Phase 1 



10/3- 


10/7 




4 


94.5 


40 


3,780 


10/4- 


10/5 






20 


9 


180 


10/5- 


10/7 




2 




9 


423 


10/5- 


10/6 








27 


567 


10/6- 


10/7 








27 


567 


10/7- 


10/8 








40 


680 


10/7- 


10/8 








27 


459 


10/8- 


10/11 




1 


73.5 


9 


662 


10/9- 


10/10 






23.5 


14 


329 


10/11 








8 


27 


216 


Subtotal 




7 






7,863 


haae 


2 












11/4 




3 




10.5 


40 


420 


11/4 




2 




10 


27 


270 


11/5 




2 




10.5 


40 


420 


11/5 




1 




8.5 


40 


340 


11/5 




3 




6 


27 


162 


11/6 




2 


2 


9.5 


27 


257 


11/6 




3 




9.5 


40 


380 


11/6 




7 




7 


40 


280 


11/6 




8 




7 


40 


280 


11/7 




7 


1 


10 


40 


400 


11/7 




9 




10 


40 


400 


11/7 




2 


1 


8.5 


27 


230 


11/7 




3 




8.5 


40 


340 


11/8 




2 


1 


9.5 


27 


257 


11/8 




3 




9.5 


40 


380 


11/8 




7 


2 


8.5 


40 


340 


11/8 




10 




8.5 


40 


340 


11/9 




3 




10.5 


40 


420 


11/9 




11 




10.5 


40 


420 



39 



Table 2. — Continued. 



Date 


Net 


Number 


Duration 


Length of 


Netting effort 


1983 


location 


captured 


in hour 6 


net (m) 


(meter-hours) 


11/9 


2 




9 


27 


243 


11/9 






9 


40 


360 


11/10 






9.5 


40 


380 


11/10 




2 


9.5 


40 


380 


11/10 






8.5 


46 


391 


11/10 


11 




8.5 


40 


340 


11/11 






10 


40 


400 


11/11 




2 


10 


40 


400 


11/11 


12 




10 


40 


400 


11/11 






10 


18 


180 


11/11 




1 


5 


27 


135 


11/12 




1 


10.5 


18 


189 


11/12 






10.5 


40 


420 


11/12 






10.5 


40 


420 


11/12 


12 




10.5 


40 


420 


11/12 






8.5 


77 


230 


11/13 






8.5 


40 


340 


11/13 




1 


8.5 


27 


230 


11/13 






8.5 


40 


340 


11/13 


13 




8.5 


18 


153 


11/13 


14 




8.5 


40 


340 


11/14 






10.5 


40 


420 


11/14 






10.5 


40 


420 


11/14 


11 




10.5 


18 


189 


11/14 






10.5 


27 


284 


11/14 


15 




10.5 


40 


420 


11/15 






10.5 


40 


420 


11/15 






10.5 


40 


420 


11/15 






10.5 


27 


284 


11/15 


16 




10.5 


40 


420 


11/15 


17 




10.5 


18 


189 


11/16 


2 




6.5 


40 


260 


Subtotal 




14 






16,753 


Total 




21 






24,616 



Table 3. — Tag numbers, capture sites, and straight carapace 
lengths of green turtles. 





Date 


Time of 


Net 


Straight carapace 


length 


(cm) 


Tag 


Midline to posterior Mid 


ine to 


No. 1 


1983 


capture 


location 


of postcentral 


of notch 


7451-55 


10/4 


0700 


1 


100.1 




99.8 


7485-89 


11/6 


1500 


2 


95.9 




94.9 


7565-69 


11/13 


1730 


2 


92.5 




92.9 


7461-65 


10/5 


1330 


1 


90.9 




89.6 


7490-94 


11/7 


1600 


7 


89.7 




89.5 


7500-04 


11/8 


1600 


2 


89.5 




88.7 


7512-16 


11/10 


0930 


2 


89.0 




88.8 


7468-72 


10/6 


1730 


1 


88.2 




— 


7456-60 


10/4 


1300 


1 


87.4 




86.9 


7560-64 


11/12 


1600 


2 


87.0 




— 


7473-75 


10/7 


1230 


3 


84.0 




83.6 


7521-25 


11/11 


1500 


7 


83.7 




83.6 


7517-20 


11/10 


1600 


2 


83.3 




83.3 


7495-99 


11/7 


1230 


2 


82.9 




82.1 


7555-59 


11/11 


1700 


1 


79.1 




78.3 


7476-80 


10/11 


1500 


3 


77.2 




76.5 


7509-11 


11/8 


1700 


7 


75.6 




75.2 


7505-08 


11/8 


1700 


7 


75.2 




74.5 


7551-54 


11/11 


1500 


7 


75.2 




74.8 


7481-84 


11/6 


1130 


2 


72.8 




72.1 


7466-67 


10/5 


1830 


3 


57.4 







'Tag series used at Johnston Atoll— 7451-7525 and 7551-7569. 

Tag inscription reads: WRITE HIMB 
UNIVERSITY 
HAWAII, 96744 



40 



Table 4. — Body measurements and weights of green turtles. 





Carapace 


length 


Carapace 


width 


Plastron 


Tail 


Bead 
















Tag 


Straight 


Curved 


Straight 


Curved 


length 


length 


width 


Weight 


Ho. 


(cm) 


(cm) 


(en) 


(cm) 


(ex) 


(cm) 


(cm) 


(kg) 


7451 


100.1 


107.8 


83.2 


105.5 


81.1 


25.6 


13.5 


.. 


7485 


95.9 


103.7 


72.5 


94.8 


80.2 


26.5 


12.8 


151.4 


7565 


92.5 


96.7 


72.1 


96.0 


79.3 


26.4 


12.6 


141.8 


7461 


90.9 


97.5 


69.8 


90.6 


73.1 


20.2 


12.1 


— 


7590 


89.7 


94.5 


71.6 


88.5 


67.5 


'41.7 


11.6 


115.9 


7 500 


89.5 


96.0 


68.8 


85.5 


72.4 


25.2 


11.9 


— 


7512 


89.0 


95.0 


67.3 


86.5 


70.8 


'54.6 


12.3 


114.1 


7468 


88.2 


92.6 


71.6 


86.0 


70.9 


'44.0 


12.1 


— 


7456 


87.4 


95.4 


70.4 


94.6 


72.8 


21.0 


11.2 


— 


7560 


87.0 


92.6 


67.4 


85.3 


70.6 


21.2 


12.0 


108.6 


7473 


84.0 


90.5 


65.4 


94.5 


70.0 


21.0 


12.0 


96.4 


7521 


83.7 


92.2 


62.2 


80.0 


67.2 




11.4 


85.9 


7517 


83.3 


87.2 


64.6 


82.0 


68.2 




— 


84.5 


7495 


82.9 


89.2 


65.4 


80.6 


68.3 


18.2 


11.3 


87.7 


7555 


79.1 


84.5 


61.6 


79.0 


64.0 


18.2 


10.7 


74.1 


7476 


77.2 


84.5 


59.6 


82.0 


61.2 


14.5 


10.2 


67.7 


7509 


75.6 


82.0 


57.0 


77.8 


— 


— 


10.3 


68.7 


7505 


75.2 


81.5 


57.8 


75.5 


60.4 


17.0 


9.9 


65.0 


7551 


75.2 


80.0 


61.2 


77.6 


59.8 


14.5 


10.9 


62.3 


7481 


72.8 


79.1 


58.1 


76.2 


59.3 


12.5 


9.7 


63.6 


7466 


57.4 


63.0 


46.3 


58.0 


44.0 


9.5 


7.1 


— 



'Adult sale. 
Short deformed tail — no indication of being a sale. 
Adult male — tail partly amputated. 



Table 5. — Front flipper neasureaents and scale counta of green turtles. 















Poatocular 






Flipper 


width* 


Fl ipper 


acalea 


aca 


lea 


Tag 
No. 


Straight carapace 
length (cm) 














Left 


Right 


Left 


Right 


Left 


Right 


7451 


100.1 


14.1 


14.4 


7 


6 


3 


4 


7485 


95.9 


14.9 


15.3 


6 


6 


4 


4 


7565 


92.5 


14.0 


14.3 


6 


6 


4 


4 


7461 


90.9 


13.2 


13.4 


6 


6 


4 


4 


7490 


89.7 


13.6 


13.5 


6 


6 


4 


4 


7500 


89.5 


14.7 


14.6 


6 


6 


4 


4 


7512 


89.0 


15.4 


15.0 


6 


6 


4 


4 


7468 


88.2 


14.2 


14.5 


6 


6 


5 


4 


7456 


87.4 


12.4 


12.6 


5 


5 


4 


4 


7 560 


87.0 


13.2 


13.2 


6 


6 


4 


4 


7473 


84.0 


13.4 


13.2 


6 


6 


4 


4 


7521 


83.7 


12.6 


12.8 


6 


6 


4 


4 


7517 


83.3 


14.7 


14.6 


6 


6 


4 


4 


7495 


82.9 


12.0 


12.0 


6 


6 


4 


4 


7555 


79.1 


13.1 


13.1 


6 


6 


5 


4 


7476 


77.2 


11.9 


11.0 


6 


6 


4 


4 


7509 


75.6 


12.4 


12.0 


6 


6 


4 


5 


7505 


75.2 


11.0 


11.2 


6 


6 


4 


4 


7551 


75.2 


12.0 


11.2 


6 


6 


4 


4 


7481 


72.8 


11.2 


10.3 


6 


6 


4 


4 


7466 


57.4 


— 


— 


6 


6 


4 


4 



'Straight line measurement taken from the anterior distal edge of the 

clav scale to the scale located directly across on the flipper's trailing 

edge (usually scale Ho. 6 counting proximal to distal along the trailing 
edge). 



41 



Table 6. — Sex determination of green turtles 



Tag 


Straight carapace 


Tail length 1 


Testoaterone 




No. 


length (cm) 


(cm) 


level 2 


Sex 


7451 


100.1 


25.6 





Female 


7485 


95.9 


26.5 


<11.1 


Female 


7565 


92.5 


26.4 


— 


Female 


7461 


90.9 


20.2 


— 


Female 


7490 


89.7 


41.7 


11,248.6 


Male 


7500 


89.5 


25.2 


<11.1 


Female 


7512 


89.0 


54.6 


561.6 


Male 


7468 


88.2 


44.0 


— 


Male 


7456 


87.4 


21.0 


— 


Female 


7560 


87.0 


21.1 


13.3 


Female 


7473 


84.0 


21.0 


— 


Female 


7521 


83.7 




13.9 


Female 


7517 


83.3 




2,324.9 


Male 


7495 


82.9 


18.2 


13.7 


Female 


7555 


79.1 


18.2 


418.7 


Male 


7476 


77.2 


14.5 


— 


— 


7509 


75.6 


— 


— 


— 


7505 


75.2 


17.0 


<11.1 


Female 


7551 


75.2 


14.5 


— 


— 


7481 


72.8 


12.5 


26.1 


Female 


7466 


57.4 


9.5 


<11.1 


Female 


Total 








13 Femalea 
5 males 



Straight line measurement from the posterior midline edge of the 
plastron to the tip of the extended tail. 

2 Testosterone level in the blood in picograms per milliliter (10~ 12 g/ml). 
3 Short deformed tail — no indication of being a male. 
"*Adult male — tail partly amputated. 



Table 7. — Identification of stomach contents sampled from green turtles. 



Tag Straight carapace Capture site 
No. length (cm) Sex (net location) 



Contents 



7451 100.1 



Oscillatoria sp. 
(trace filaments) 



7565 



7461 
7512 



7473 
7495 



7476 
7466 



90.9 
89.0 



84.0 



77.2 
57.4 



Bryopflia pennata var. 

secunda 
Oscillatoria sp. 
Unidentified amphipod. 1 
Pyxidula sp. (diatoms- 

trace) 

Zonarig sp. (trace- 
filament) 

Caulerpa racemosa var. 
uvifera 

ClimocoBphenia sp. 
(diatoms-trace ) 

Oscillatoria sp. (trace- 
filament) 

Unidentified fibers- 
trace 

Climacospbenia sp. 

(diatom-trace) 
Unidentified filamentous 

bacteria 

£• Penngta var. se cunda 

,B. pennata var. secunda 
Polysiphonia sp. 
(fragment) 



Probably originated from the esophagus. 



42 



Table 8. — Identification of mouth and fecal contents 
sampled from green turtles. 

Tag Straight carapace Capture site 

No. length (cm) Sex (net location) Contents 

Mouth contents 

7485 95.9 Female 2 Caulerpa racemosa var. 

macrophyaa 
Acrochaetium sp. 

(epiphytic on Caulerpa ) 

7565 92.5 Female 2 Brvopsia pennata 

Oscillatoria sp. 

filaments 
Pyridicula sp. (diatoms) 
Unidentified amphipod 

7560 87.0 Female 2 Bryopsis pennata var. 

secunda 
Pyxidicula sp. (diatoms) 

7555 79.1 Male 1 Caulerpa racemosa var. 

macrophyaa 
Onidentified filamentous 

bacteria 
Unidentified blue-green 

»lg«e 

7481 72.8 Female 2 Bryopsis pennata 

Fecal contents 

7509 75.6 — ' Caulerpa racemosa vsr. 

uvifera (75X) 
Bryopsis pennata (251) 



Not determined. 



Table 9. — Algae collected during diving surveys vith scuba. 

Dive Date 

No. 1983 Location Algae collected 

10 10/10 J Bryopsis pennata var. secunda (Harvey) 

Collins and Harvey 
Caulerpa racemosa var. macrophyaa 

(Rutting) Taylor 
C_. racemosa var. uvifera (Turner) Weber 

von Bosse 
C. serrulata (Fbrskal) J. Ag. 

11 10/11 t C. serrulata f. anguata (Weber von 

Bosse) Taylor 
C. serrulata (Forskal) J. Ag. 
Dictyota friabilis Setchell (epiphytic 

on Caulerpa ) 

13 11/4 M Gelidium pus ilium 

Ceraminium ap. (trace) 

14 11/5 N Caulerpa serrulata 

Avrainvillea lacerata 
Hydrocoleum lyngbyaceum 
Zonaria ap. 
Polysiphonia sp. (trace) 



43 



Table 10. — Percent nutrient composition of principal food 
sources used by green turtles. 1 

Acid detergent fiber 2 

Neutral 

Dry Crude Ether detergent Permanganic 
Algae matter protein 3 extract Ash fiber 2 lignin Cellulose 

Brvopaia pennata 7.0 23.8 2.0 38.2 25.6 2.7 7.9 
var. secunda 
(from foraging 
site) 

B. pennata 7.1 25.7 2.6 33.8 27.8 3.3 11.3 
var. secunda 

(detached) 

Caulerpa racemose 3.9 8.0 0.7 61.4 24.6 6.3 6.3 
var. uvifera 

C. racemosa 3.7 9.1 0.9 63.8 25.5 6.5 7.5 
var. macrophvsa 

Reported on a dry matter basis as determined by the "proximate analysis" 
method commonly used for terrestrial animal forage. 

Present in benthic algae as a complex polysacchride; not true lignin or 
cellulose as found in terrestrial plants. 
'Nitrogen x 6.25. 



Table 11. — Mineral composition of principal food sources used by 
green turtles. 1 



Algae Ca P K Hg Na Fe Cu Mn Zn 

X ppm 



Brvopsis pennata 2.00 0.27 0.94 1.06 11.10 110 8 19 57 
var. secunda 
(foraging site) 

B. pennata var. 1.82 0.23 0.93 0.95 10.00 88 10 17 66 
secunda 

(detached) 

Caulerpa racemosa 2.45 0.10 1.07 0.41 19.36 90 9 9 76 
var. uvifera 

C. racemosa 0.87 0.10 1.28 0.31 22.00 2,558 11 29 81 
var. macrophvaa 



•Dry matter basis. Ca ■= Calcium; P - Phoaphorus; K - Potassium; 
Hg - Magnesium; Na - Sodium; Fe - Iron; Cu - Copper; Mn ■= Manganese; 
Zn - Zinc. 



44 



Table 12. — Identification of epizoites sampled from green turtles. 

Tag Straight carapace 

No. Sex length (cm) Epizoites 

7485 Female 95.9 Acrochaetium ap. 

Polyaiphonia tsudana 
Lyn yb ya semiplena 
Unidentified roundworms 
Unidentified amphipoda 
Unidentified black "mitea" 

7565 Female 92.5 Acrochaetium ap. 

Polyaiphonia taudana 
L. aemiplena 
Sphacelaria tribuloidea 
Uroaopora a p. 
Pilinia ap. f 
Unidentified foraminif era 

7512 Male 89.0 Acrochaetium ap. 

Polyaiphonia taudana 
L. aemiplena 
Uroapora ap. 
Pilinia ap. l 
Chadophora ap. (trace) 
Dermocarpa aphaerica (epiphitic 
on Chadophora ap.) 

7495 Female 82.9 Same aa tag No. 7485 

7431 Female 72.8 Polyaiphonia taudana 

L. aemiplena 
Uroapora ap. 
Pilinia ap. 



'Possibly Pilinia rimoaa Putting, which may be a new record for the 
tropical Pacific. 



Table 13. — Red blood cell chol ineateraae valuea for nine green turtlea. 



Tag 
No. 


Straight carapace 
length (en) 


Weight 
(kg) 


Net 
location 


Sex 1 


Cholineiteraae 2 


7485 


95.9 


151.4 


2 


Fenale 


0.15 
0.16 


7490 


89.7 


115.9 


7 


Male 


0.27 
0.27 


7500 


89.5 


- 


2 


Fenale 


0.13 


7512 


89.0 


114.1 


2 


Male 


0.23 


7517 


83.3 


84.5 


2 


Male 


0.29 
0.30 


7495 


82.9 


87.7 


2 


Fenale 


0.14 
0.13 


7555 


79.1 


74.1 


1 


Hale 


0.32 
0.34 


7505 


75.2 


65.0 


7 


Fenale 


0.26 


7481 


72.8 


63.6 


2 


Fenale 


0.28 
0.31 



*Baaed on teatoaterone level. 

decrease in pB unita (pR/h) using a 0.2 ml aliquot. 



45 



I6' , 44'N 




169° 32 W 



Figure 1. — Location of turtle nets. 



46 



16° 44 N 




169° 33' W 




Figure 2. — Central location of 26 diving surveys with scuba. The bottom 
areas actually covered extend out in concentric circles from each loca- 
tion and overlap considerably for many dives. 



ATOLL RESEARCH BULLETIN 
No. 286 



ENVIRONMENTAL SURVEY OF MATAIVA ATOLL, 
TUAMOTU ARCHIPELAGO FRENCH POLYNESIA 



By 
B. Delesalle and Colleagues 



Issued By 
THE SMITHSONIAN INSTITUTION 
Washington, D- C«, U.S.A. 
May 1985 



CONTENTS 

Page 

INTRODUCTION 1 

PRESENTING MATAIVA ATOLL 3 

Geography 3 

Background and population 3 

Economy 3 

GEOLOGICAL SETTING OF MATAIVA ATOLL 4 

Geomorphology of Mataiva 4 

Outer reefal and lagoonal sediments 5 
Sequence of the main geological events at 

Mataiva atoll 6 

HYDROLOGICAL ENVIRONMENT 7 

Currents 7 

Water level 7 

Temperature and salinity 8 

Turbidity and light penetration 8 

Dissolved oxygen 8 

Nutrients 9 

PRIMARY PRODUCERS OF MATAIVA LAGOON 9 

Phytoplankton 9 

Benthic macroflora 10 

MATAIVA LAGOON FAUNA ■ 11 

Zooplankton 11 

Corals 12 

Molluscs 12 

Crustacean fauna 13 

Other marine invertebrates 14 

Fishes 15 

CONCLUSION 16 

REFERENCES 17 

TABLES AND ILLUSTRATIONS 19 



In addition to the author's own research results, this paper in- 
cludes contributions by many colleagues, from unpublished reports, 
theses in progress and other data, which are here gratefully acknow- 
ledged. These collaborators and their affiliations are listed below. 
All are also attached to the Centre de l'Environnement in Moorea(l). 

J. Bell^ 3 ), F. Bourrouilh-Le Jan< 4 ) , J. de Vaugelas^ 5 ) , C. Gabriel, 

R. Galzin^ 2 ) , M. Harmelin^ 6 ), L. Montaggioni^ 7 ) , M. Monteforte^ 2 ) , 

0. 0dinetz (2) , C. Payri (2) , M. Pichon (8) , J.P. Renon (9) , M. Ricard (10) : 

G. Richard( 2 ), B. Salvat^ 2 ). 



(1) Centre de l'Environnement Antenne Museum-EPHE, BP 12 Moorea, 
Polynesie Francaise 

(2) Laboratoire de Biologie Marine et Malacologie, EPHE, 55 rue de 
Buffon, 75005 Paris - France 

(3) School of biological Sciences, Macquarie University, North Ryde, 
N.S.W. 2113, Australie 

(4) Laboratoire de Geodynamique, Universite de Pau, 64000 Pau - France 

(5) Laboratoire de Biologie et d'Ecologie marines, Universite de Nice, 
28 Avenue de Valrose, 06034 Nice Cededex - France 

(6) Station marine d'Endoume, Rue de la batterie des Lions, 
13007 Marseille - France 

(7) Universite Francaise de l'Ocean Indien.BP 5, 97490 Sainte Clotilde 
lie de la Reunion - France 

(8) James Cook University, Queensland 4811, Australia 

(9) Laboratoire d'Ecologie animale, Universite d'Orleans, 45100 
Orleans - France 

(10) Laboratoire de Cryptogamie, Museum National d'Histoire Naturelle, 
12 rue de Buffon, 75005 Paris - France 



ENVIRONMENTAL SURVEY OF MATAIVA ATOLL, 
TUAMOTU ARCHIPELAGO FRENCH POLYNESIA 

By 
B- Delesalle and Colleagues * 

INTRODUCTION 

Mataiva Atoll, one of 84 in French Polynesia, is a small coral 
island at the western edge of the Tuamotu Archipelago. This atoll 
occupies a particular place among the French Polynesian atolls since the 
discovery beneath the lagoon sediments of deposits of phosphates soon to 
be exploited. In order to estimate the environmental effects of such 
exploitation and plan a management scheme, numerous studies have been 
carried out since 1978 by many scientific organizations: Antenne du 
Museum National et de l'Ecole Pratique des Hautes Etudes en Polynesie 
Franchise, Office de la Recherche Scientifique et Technique d'Outre-Mer 
(ORSTOM), B.C. Research, Institut de Recherches Medicales Louis Malarde 
(IRMLM), Centre National pour l'Exploitation des Oceans (CNEXO) and 
Commissariat a I'Energie Atomique (CEA-LESE). 

Because of the unusual and interesting results of these impact 
studies, the Antenne Museum-E.P.H.E. has, since 1981, continued the 
scientific survey of Mataiva Atoll. The geology and geomorphology of the 
atoll, the hydrological characteristics of lagoon waters and the 
abundance and diversity of the marine flora and fauna have been 
investigated by about 20 scientists. Several contributions, including 
for example studies of fishes, crustaceans and phy top lank ton, are 
extracted from doctoral manuscripts. 

Mataiva atoll has been chosen as the site of a post-Congress field 
trip of the Fifth International Coral Reef Congress, to be held in Tahiti 
in June 1985. 



* Centre de 1' Environnement Antenne Museum-EPHE, BP 12 Moorea, Polynesie 
Francaise 

Laboratoire de Biologie Marine et Malacologie, EPHE, 55 rue de Buff on, 
75005 Paris, France 



PRESENTING MATAIVA ATOLL 
Geography 



o 



,o 



Mataiva atoll is located 300 km north of Tahiti, 14 55' lat. S and 
148~ 36' long. W. It is the westernmost atoll in the Tuamotu 
Archipelago. This small island, 10 by 5 km, is distinguished by its 
unusual morphology: a wide atoll rim, almost continuous, allowing only 
limited oceanic exchanges, and a reticulated lagoon divided into numerous 
basins by a network of slightly submerged partitions (Figure 1). 

The climate is not unlike that elsewhere in Polynesia, i.e. tropical, 
hot and humid, with 2 seasons: one, relatively dry and cool (24-27° C) 
from April to September, and the other, hot and rainy (28-30°C) from 
October to March. Since the atoll has little relief, rainfall rarely 
exceeds 2.5 m per year, and the amount of sunshine is high (2500 hours 
per year). Storms and cyclones are rare with the exception of 1983, when 
3 cyclones hit Mataiva. Trade winds are dominant, blowing from the 
eastern sector at an average speed of 7 to 10 m/s. The swell is 
generally less than 4 m, coming mainly from the South. 

Background and population 

Europeans discovered Mataiva in 1819: it was then named Lazareff 
Island by Bellingshausen. The numerous archeological discoveries bear 
witness to a more or less permanent pre-European settlement. Mataiva 
then belonged to the 8 independent kingdoms of the western Tuamotu, which 
formed the cultural and linguistic area of Mihiroa. However, this island 
was often uninhabited, particularly after its people were massacred in 
the 18th century by the inhabitants of Anaa atoll. 

The present population on Mataiva is of recent origin. Around 1940, 
representatives from various families of the neighbouring atoll, Tikehau, 
who were exploiting Mataiva's copra, decided to settle there. The atoll 
was declared autonomous in 1950 and in 1971 was associated with the 
community of Rangiroa (Jarrige et al., 1978). 

In 1983, the population of Mataiva numbered 183. It is a population 
of predominantly younger generation (55.8% are under 20 years old), 
slightly unbalanced (63.4% being female) and very fluctuating because of 
its proximity to Tahiti. The relationship, with Tikehau and Makatea, to 
the larger community of Rangiroa, explains the frequent population surges 
to these islands, whether temporary or permanent. For example, of the 
172 persons established on Mataiva recorded in a survey taken in 1964 
only 41 were still there in 1983. 

Economy 

Fishing and copra production are the two traditional means of 
support on the Tuamotu Atolls. These two natural resources are enough to 
insure the subsistence of the inhabitants and to allow the surplus to be 
commercially exploited. Being close to Tahiti, the main market, Mataiva 



glories in its privileged position; moreover, it will be able to add the 
yield from the exploitation of phosphates to its future revenue. 

Fishing. . On atolls generally, the sea provides about 70% of the 
proteins consumed, mainly through fish, but also sea food (shell fish and 
crustaceans) collected along the reef, and turtles which are considered a 
great delicacy. Fishing techniques vary: spears ("pupuhi"), rods 
("aira"), lines or nets are used, but commercial fishes are above all 
collected in fish traps, then sold to passing schooners. These fish 
traps are located in the pass and next to the hoa (ocean-lagoon channels) 
and passively catch reef fishes. The amount of the catch varies greatly 
throughout the year and provides around 45 tons annually. 

Copra. An essential atoll resource, copra occupies a large majority 
of the population throughout the whole year: the upkeep of the coconut 
grove, harvesting and drying of the nuts and the export to Tahiti. Due 
to the size of its emerged reef rim, Mataiva is an important producer 
with an annual harvest of nearly 450 tons, i.e. about 5% of the total 
production in the Tuamotu Archipelago. Unfortunately, this production 
was completely eliminated as a result of the 1983 cyclones and for at 
least two years, the main activity has been devoted to the restoration of 
the devastated coconut plantations. 

Phosphates* The phosphate deposit represents the future--al though at 
the moment hypothetical—wealth of Mataiva. It extends for approximately 
5 km and represents 15 millions tons of ore, 10 of which can be 
extracted. Its exploitation, forecast to last 10 to 15 years, will 
necessitate the employment of about 200 people. Such a project wilL 
result in complete upheaval in the atoll's morphology and the way of life 
for its inhabitants, but should not do so to the detriment of its 
traditional resources. One hopes that Mataiva will not experience the 
same fate as its neighbouring island, Makatea, which was practically 
deserted following the exploitation of its deposits. 

GEOLOGICAL SETTING OF MATAIVA ATOLL 

The morphological and sedimentological characteristics of Mataiva 
atoll are rather different from those encountered in most other atolls of 
French Polynesia. These singularities appear to be a consequence of 
unusual and late geological events which caused, among other things, the 
formation of a reticulated lagoon and deposits of exploitable phosphates. 
In particular, tectonic uplift of the NW Tuamotu atolls presumably 
resulted from the loading effects of the nearby volcanic complex. A 
crustal moat has developed peripheral to Tahiti, Moorea and Mehetia 
volcanoes. Beyond the outer edge of the moat, flexuring (buckling) has 
developed an arch which has been uplifted by about 10 meters (Pirazzoli 
and Montaggioni, in press). 

Geomorphology of Mataiva 

When approaching Mataiva by plane, one is immediately struck by 2 
morphological characters: a wide reef rim and a partitioned lagoon. 



The emerged reef ring is 200 to 1500 m wide, i.e. somewhat wider than 
found in the other atolls of the Tuaraotu Archipelago. This reef rim is 
almost continuous, broken only by some channels (hoa) on its southern 
coast and a small atypical pass in the NW. The N and E coastlines form a 
single islet (motu). 

A topographic cross-section (Figure 2) allows a proper understanding 
of this atoll's morphology. It reveals a marked dissymmetry between the 
N end S coasts of the atoll — the N and NE coasts are narrow (200-500 m) 
and of relatively high altitude (+ 6 m) , whereas the southern coasts are 
lower and wider (1000-1500 m) . In the same way, this asymmetry continues 
to the inner slope, which is wider (300 m) on the N coast than on the 
southern one (50 m) . This dissymmetry is easily accounted for by the 
violent storms coming from the N wit-h the resultant accumulation of 
storm ridges on the N coast, while the lagoon muds are transported 
southwards where they contribute to the supratidal accretion of the 
southern motu. This high energy sedimentation contrasts with that which 
prevails in a "normal" meteorological situation, where the influence of 
the S E tradewinds favours rather the widening of the northern inner 
slope. 

The outer reef, of variable width (150-500 m) , is characterised by 2 
fundamental features : 

- The presence of a reef flat flagstone, more or less eroded, located 
below and behind the algal ridge, where biodetritic sedimentation is weak 
or non-existent. 

- The presence of an algal ridge in the process of chemical erosion; 
remnants of a raised, fossil ridge appear in some places; crusts of 
living coralline algae are spotty. Construction is quite inactive and 
the relief is in the process of degradation. 

The lagoon morphology is the most original characteristic of Mataiva 
atoll. It is a reticulated or partitioned lagoon, made up of about 70 
pools of varying sizes (100 m to over 2 km), with an average depth of 8 
m. The shallows which separate these pools are under 0.1-0.8 m of water 
and form a network over 200 km long. 

This sort of reticulated lagoon seems to be the only one in French 
Polynesia but its morphology is reminiscent of certain reef zones of Bora 
Bora (Society Archipelago) or of Ponape (Caroline Islands), which also 
consist of such lagoon pools each isolated from the others. Surveys 
conducted in the lagoon reveal that this reticulation seems directly 
related to paleotopography, most likely a pre-Pleistocene surface, buried 
under 10-15 m of present-Holocene sediments. This paleotopography 
influences the present topography and makes up the framework of the 
atoll's general morphology. However, Holocene coral remnants located 
30-40 cm above mean lagoon level, determine the morphology of the lagoon 
coast and form some small islets in the eastern and southwestern parts of 
the lagoon. 

Outer reefal and lagoonal sediments 

Grain-size analyses of bottom surface samples have only been made on 
the fraction larger than 40 um. Two textural parameters (mean size, Mz 



and sorting, So) have been graphically determined from the cumulative 
frequency curve. There appears to be a sharp difference between the 
reefal and lagoonal sediments (Figure 3). On the reef rim, the sediments 
are medium to coarse sands or granules (Mz= 0.67-3.45 mm), usually well 
sorted (So = 0.76-0.91), except in the western sector (So= 2.18). In all 
cases, fractions smaller than 0.5 mm do not exceed 25%. On the contrary, 
sediments in the lagoon are fine or very fine sands, to muddy sands 
(10-50% of fractions < 40 u_m), or to sandy silts (over 50% < 40 urn). The 
sorting is usually good (So = 0.28-1.48). Higher percentages of silts 
have been found on the shoals sides or at the bottom of the basins, 
mainly in the N and E sectors of the atoll (50-80% of silts). 

The composition of sediments reveals 2 main sources, corals and 
Corallinaceae. Green calcareous algae ( Halimeda ) , Forarainifera and 
molluscs can also make up an important part of the sediment. Crustacea, 
serpulids, sponges and Bryozoa are also found. Figure 3 shows the grain 
size and composition of sediments in different parts of the outer reef 
and lagoon. The important features to emphasise are the abundance of 
Foraminifera on the southern outer reef, the great variation of 
percentages of Halimeda and Foraminifera in the lagoon, and the better 
representation of Corallinaceae and molluscs in the lagoon than on the 
outer reef. 

Sequence of the main geological events at Mataiva atoll 

Geological investigations by subsurface drilling showed the dominant 
influence of the pre-Holocene surfaces on the topography of Mataiva's 
reef structures. By reference to the geological history of other French 
Polynesian atolls and particularly that of Makatea (Montaggioni, 1985), 
the geological history of Mataiva can be summed up as follows : 

During the Mio-Pliocene(T) a platform-like reef developed. This old 
reef is at present emerging along the N coast in the upper part of the 
ocean side beach. In the central and southern parts of the atoll, it is 
buried under a layer (a few to 30 m deep) of Holocene deposits. During 
several spells of emergence, this old reef underwent severe meteoritic 
solution and partial dolomitization. 

During a later stage (Pliocene ?), the cavities of the karst thus 
formed were filled up with a phosphate deposit. The contrast in facies 
between the rocks of the old reef and the deposit demonstrates the 
totally different conditions of deposition for the phosphate. However, 
at present it is not possible to offer a reliable reconstruction of the 
deposition environment of the phosphates. Three alternatives may be 
proposed for their origin: on weakly consolidated carbonate rocks (high 
residual porosity), phosphorites may have been formed by (a) alteration 
and diagenesis of sea birds excrements and bones, (b) alteration of 
marine organic matter, deposited under low oxygen conditions, (c) 
post-depositional alteration of drift volcanic material that accumulated 
as soils. 

Subsequent to phosphatogenesis , Mataiva Island underwent a slight 
tectonic uplift which appears to be confirmed by the absence of deposits 
linked to the 120,000 B.P. high sea level, which is found at Moruroa 



atoll between 7 and 11 ra below the Holocene deposits (Buigues, 1983). 

The island was later submerged and thus the phosphate deposits and 
the rest of the old reef were recolonised by coral growth during the 
Holocene transgression, about 6,000 years ago. 

Lastly, between 5,000 and 2,000 B.P., coral growth developed at 
Mataiva at a level slightly higher than the present one. The relicts of 
algal ridges found on the outer reef (2,200 + 130 yrs) and the Porites 
colonies of the lagoon motu (5,210 + 130 yrs) are at elevations similar 
to the emerged beachrocks,in which lithif ication patterns indicate an 
exposure patterns to vadose diagenetic environments during their 
formation (Montaggioni and Pirazzoli, 1984). 

HYDROLOGICAL ENVIRONMENT 

The hydrological characteristics of Mataiva atoll have a direct 
connection with 2 main morphological features: (1) a lagoon of reduced 
size and depth, whose shoals further restrict the volume (2) limited but 
still existing relations with the ocean through hoa and pass. Conse- 
quently, the physicocheraical characteristics of the lagoon waters, if 
they show a certain spatial homogeneity, present an enormous temporal 
variability, principally dictated by climatic conditions (Delesalle, 
1982). 

Currents 

The hoa's position, close together on the S side of the atoll and 
facing the dominant swells, in relation to that of the pass on the 
opposite sheltered coast, brings about a general circulation of the 
waters from the S towards the NW. Oceanic waters enter the lagoon 
through the hoa, whilst, in the pass, the current is most often outgoing 
(Figure 4). 

In the lagoon, the partitions slow down considerably the water's 
circulation, and only the wind-induced currents are measurable. However, 
because of the larger size of the basins along the northern and southern 
coasts, the flow of water may preferentially follow them, and avoid the 
atoll's centre. In this case, the eastern part of the lagoon, isolated 
by a transversal string of islets, appears more confined. 

Water level 

The small amount of ocean-lagoon exchanges, added to the relatively 
reduced volume of the lagoon, explains how the water level can undergo 
considerable variations. An example of such variations, recorded daily 
since 1979 next to the pass, is given in Figure 5. Their usual amplitude 
is about a metre, but can reach or surpass 1.5 m when a very strong swell 
occurs. On the other hand, these variations are very rapid, a rise of 40 
cm in 24 hours is not unusual. Considering the area of the lagoon (2500 
ha), such an increase corresponds to an entry of water of about 10 m , 
i.e. 1/10 of the lagoon volume. The lowering of this level is more 
gradual, around 10 cm in 24 h, but can last for periods of 7 to 10 days, 
and thus reach such a low level that the coral colonies on the top of the 
partitions emerge. 



8 

The general evolution of the level over 5 years shows a certain 
predominance of lower levels between July and December, but without well 
defined cycles. 

The variations in the lagoon level show that Mataiva cannot be called 
a closed atoll. The existence of a pass, although atypical and shallow, 
guarantees permanent oceanic exchanges. 

Temperature and salinity 

Water temperatures in the lagoon do not vary much and generally 
follow the air temperatures with a maximum (29 -31 C) in the rainy season 
and a minimum (25 -27 C) during the dry season. If there is a slight 
warming up of the surface waters during the day, still a marked thermic 
stratification does not occur as could be expected in the absence of 
circulation. 

This homogeneity of the water column is confirmed by salinity levels 
which differ little between surface and bottom. Temporal variations in 
salinity are more marked. In fact, if the lagoon waters near the hoa show 
salinity values usually close to those of the ocean (36 g/1), in the rest 
of the lagoon drops in salinity (30.77 to 33.89 g/1 in April 1981) or 
increases (36.56 to 37.64 g/1 in October 1983) can be observed. 

If climatic conditions (heavy rains or long periods of dryness) 
directly influence the salinity of lagoon waters, the phreatic 
freshwater, held in the atoll foundation, also plays a role: the low 
salinity levels observed in April 1981 were measured after a month 
without noticeable precipitation while over 500 mm of rain had been 
recorded the month before. 

Turbidity and light penetration 

The lagoon waters of Mataiva always have a more or less pronounced 
milky appearance. Secchi disc measurements of water transparency reach 
50 m in the ocean, but do not exceed 7 m in the lagoon near the hoa and 
2.5 m near the pass. Quantum measurements indicate a quick absorption of 
light with depth (Figure 6). The attenuation coefficient deduced from 
these curves is weaker near the hoa (0.14) than towards the north of the 
lagoon (0.28); the most turbid waters (0.36) are found near the pass. 
However, the amount of suspended matter is not very great: 6.7 to 9 rag/1. 
This strong decrease in light in the water is caused by very fine, silty, 
calcareous particles. 

Dissolved oxygen 

Levels of dissolved oxygen vary very little (5.4 to 7mm/l) and 
remain close to, or higher than the level of water saturation, depending 
on temperature and salinity. Such values indicate a confined environment 
where photosynthetic organisms play a major role in water oxygenation. 



Nutrients 

Lagoon waters of atolls are usually considered to be oligotrophic 
because of the absence of continental Influences. Mataiva lagoon water, 
on the contrary, contains high and variable concentrations of dissolved 
nutrients, especially nitrates and silicates (0.10-13.99 uatg N-N03/1, 
0.3-16.5 uatg Si/1). Heterogeneity is very marked between different 
areas of the lagoon and different periods of measurements, but no 
well-defined pattern can be recognized from the distribution of the 
values. 

The confinement of the lagoon water, its lack of depth and its 
division into numerous pools, appear to be factors that allow such high 
and variable concentrations. Moreover, the migration of nutrients from 
the oceanic deep layers and the volcanic substratum through the coral 
foundation, as shown in Moruroa atoll (Rougerie et al.,1982) and in 
Takapoto atoll (Rougerie, 1983), might be an important contribution to 
the enricheraent of the waters. Although we have no information about the 
porosity of the coral foundation of Mataiva, it is probably not less than 
on Takapoto considering the geological events (uplifts, karstif ication) 
undergone by Mataiva Atoll. 



PRIMARY PRODUCERS OF MATAIVA LAGOON 

Phytoplankton 

The phytoplankton of the lagoon waters is of relatively low specific 
diversity. Six classes of algae can be identified, Diatomophyceae, 
Dinophyceae, Chlorophyceae, Cyanophyceae, Cryptophyceae and Ghrysophyceae 
(Coccolithophorideae) , but with few species (Table A). 

The diatoms dominate the phytoplanktonic flora by the diversity of 
the existing species, particularly the genera Mastogloia , typical of 
tropical seas, Nitzschia and Navicula . The scarcity of strictly 
plank tonic forms such as Rhizosolenia , Chaetoceros or Thalassiosira must 
be noted. These are found only near the hoa, where oceanic waters enter 
the lagoon. The populations are mainly made up of phytoplanktonic 
species. Dinof lagellates are also represented by species from calm and 
shallow environments: Gymnodinium , Gonyaulax and Prorocentrum (notably P.. 
ovum). The other classes, especially Cryptophyceae and Chlorophyceae, are 
not often found in the plankton of atoll lagoons. The presence, and, at 
times the abundance, of some species such as the green alga Pyramimonas 
is a peculiar characteristic of Mataiva plankton. 

3 5 
The abundance of Mataiva's phytoplankton varies from 10 to 5 10 

cells per litre. Quantitatively the small-sized phytof lagellates (10.30 

um) are often dominant while the diatoms are only locally abundant. A 

dinof lagellate Gymnodinium and a chlorophyceae Pyramimonas sometimes 

represent 90% of the cells counted. 

The presence and importance of phytof lagellates in Mataiva plankton 
are characteristic of a calm and shallow environment. However, because 
of the richness in nutrients of the waters, one might expect a high 



10 

bioraass in the plankton. On the contrary chlorophyll measurements give 
low values : 0.3 to 1 ug chl a 1 in surface waters, to 5 near the 
bottom. Never thless, the low percentage of degraded pigments (leas than 
35%) as well as the high level of primary production (23-90 mg cm d ), 
so much higher than values usually observed in atolls (Sournia et Ricard, 
1976; Delesalle et al. 1983), are indicative of a rapid turn-over in 
the phytoplanktonic populations. Such a paradox between low biomass and 
strong primary production can be partially explained by the wealth of 
zooplanktonic populations whose grazing can keep the phytoplanktonic 
biomass at a low level. 

Benthic macroflora 

The benthic macroflora in Mataiva's lagoon is characterised by a 
small number of species, of which only a few are abundant (Table B) . 
Essentially, it is a hard substratum flora. Only a sea-grass, Halophila 
cf. ovalis , forms extensive grassbeds on the sandy shoals. A filamentous 
green alga, Enteromorpha , can also on occasion form large masses on the 
north inner slope. 

Unlike high volcanic islands, where brown algae are the most abundant 
species, atoll floras are usually dominated by green and red algae. At 
Mataiva, the green algae are the most abundant and diversified (22 
species), particularly the genera Caulerpa and Halimeda . One of them, 
Halimeda ( opuntia group) is strongly developed on Mataiva, an unusual 
situation for an atoll lagoon. Among the Rhodophyceae, the crustose 
algae are more numerous in the lagoon and on the outer reef; however, 
gelidial turfs are well developed on the lagoon's dead corals and on the 
outer reef's flagstone. Cyanophyceae make up the 3rd class in the algal 
complement; if the usual extensive blue-green formations of atoll lagoon 
bottoms are not found in Mataiva, these algae do however form a discrete 
felt on dead corals and on Halophila leaves; their development is also 
important on beach edges, where they form an algal mat at times very 
thick, and in the brackish ponds of the atoll rim. 

The distribution of algae in the lagoon is fairly homogeneous. Few 
species are present simultaneously. The specific diversity and abundance 
increase considerably near the zones of water exchange with the ocean 
(hoa and pass). On the outer reef flat, the eroded flagstone is covered 
with Gelidiales turf, while the brown alga Pocockiella variegata is 
dominant on its outer zone. Near the reef front, soft algae common to 
this zone ( Microdyction , Dyctiosphaeria cavernosa , Neomeris van bosse ) 
appear, as well as crustose corallinaceae, Porolithon onkodes , 
Chevaliericrusta sp. But they never form an algal ridge typical of the 
reef front of an atoll, and their mass always remain poorly developed. 

Thus, Mataiva's marine flora is that of a closed environment, with 
few species, some of which are abundant. It is homogeneous and gradually 
changes near the ocean exchange zones. Although not a rich flora, its 
vitality is demonstrated by the presence of numerous young shoots. The 
existence of large Halophila grassbeds and the unusual development of 
Halimeda ( opuntia group) remain the distinctive characteristics of this 
flora. 



11 

MATAIVA LAGOON FAUNA 



Xooplankton 



The zooplankton of Mataiva atoll is mainly composed of typical, 
lagoonal holoplanktonic species where 6 species of Copepods, 2 species of 
Chaetognaths, 1 Appendicularia, 1 Ostracod and 1 Decapod Sergestid are 
dominant. Surprisingly, the meroplankton is extremely rare, 5% of the 
total plankton, whereas its true contribution is usually 35-65%. This 
meroplankton is made up of crustacean larvae (S tomatopods , Decapod 
Reptantia and Natantia), fish eggs and larvae, and Foraminif era. In 
other respects, while all of the Mataiva zooplankton species are recorded 
from other atolls, the absence of groups occurring in atolls open to the 
ocean, e.g. Pteropods, Salps, Doliolids and some Copepods, is noteworthy. 

The zoo_pJ.ankton bioraass is much higher than in the nearby ocean: 
300-500 mg/m in the lagoon, 10-18 mg m" in the ocean. 

The distribution of this biomass is very heterogeneous. 
Horizontally, the western and southern sectors of the atoll are about 3 
times richer than the eastern and northern areas. Vertically, a very 
marked diurnal stratification exists between the surface (80 mg m ) 
and the bottom (2 000 mg ra ) of the lagoon. This phenomenon has been 
observed in other atolls: Rangiroa (Michel, 1971), Mururoa (Michel, 
1969), Takapoto (Renon, 1977), Bikini (Johnson, 1949). 

The ocean- lagoon exchanges through the hoa and the pass are very 
important: entry of reef plankton, mainly meroplankton, and outpouring of 
lagoon plankton through the pass, in quantity about 35 times higher than 
what enters the lagoon. The lagoon thus constitutes an extremely 
productive environment, greatly enriching the nearby ocean. 

The abundance and composition of Mataiva's zooplankton are somewhat 
unusual: 

Firstly , its biomass is on the average 2 to_.3 times greater than in 
other Polynesian atoll lagoons : 300 to 500 mg ra _at Mataiva, 50 to 150 
mg m" at Moruroa (Renon, unpubl.), 47 to 61 mg m" at Takapoto (Renon, 
1977). The confinement of waters only partially explains this pheno- 
menon, since Takapoto, a closed atoll, is poorer in zooplankton. 

Secondly , the scarcity of meroplanktonic forms characteristic of 
Mataiva's zooplankton might be related to the depauperate benthic fauna, 
hence to lighter grazing and may account in part for a rich biomass. 

Finalljj, the nutritional basis of zooplankton populations remains 
difficult to define. The phytoplankton biomasses are not in equilibrium 
ith the zooplankton abundance, although the turnover rate of phyto- 
plankton appears very high; however, the seston particles, which support 
bacterial development, may be directly used by the zooplankton. 



12 

Corals 

As is often the case in closed atoll lagoons (Chevalier et 
Denizot,1979) , the specific diversity of Mataiva lagoon corals is 
especially low. Only 28 species have been recorded in the lagoon (Table 
C). The areas of maximum diversity (12-14 spp.) have been found at 
stations on the N and S coasts of the atoll. This relative variety is 
related to hydrodynamic conditions, due to the nearness of the hoa and 
the general flow of the water across the northern and southern edges of 
the lagoon. In fact, only those stations removed from the hoa and the 
pass have a low population. 

The cover rate by scleractinians is generally very poor. The 
colonies are mainly located on the tops of the lagoon partitions, along 
the edges of the pools. Porites lobata , forming microatolls on the 
shoals, and Acropora tortuosa , more generally covering the pool sides, 
are the 2 most commonly reported species. Other species, Montipora 
aequi tuber cu lata and Leptastrea purpurea , are widely distributed, but 
form a poor cover because of the sraallness of their colonies. 

The coral fauna's vitality seems to be very poor and dead colonies 
are numerous. The percentage of living corals, which is almost 30% near 
the hoa but tapers to in the eastern part of the atoll, is directly 
connected with hydrodynamic conditions. However, the size of the dead 
colonies of Porites and Acropora indicates an accidental origin for this 
condition. The combination of several environmental factors, such as a 
low water level along with much rain or intense sunshine, is likely to 
induce, in certain parts of the lagoon, hydrological conditions incom- 
patible with coral survival, thus causing massive death. 

The inhabitants of the atoll reported the occurrence of such an 
event in November 1978, repeated in November 1980: water level at 52 cm, 
extremely strong sunlight and water temperatures close to 32 C for 10 
days. The lagoon waters turned green-brown in certain parts and produced 
a nauseating odour. In May 1981, the live coral cover was low throughout 
the lagoon, exceeding 10% only near the hoa. Other measurements taken in 
October 1983 show a considerable increase in live cover. Such a 
situation is only possible as a result of oceanic inflows, allowing 
recruitment of larvae and a return to conditions favouring the growth of 
the surviving species. 

The outer reef coral fauna is much more diversified than that in the 
lagoon (Figure 10). These colonies are often small or very encrusting. 
The eroded reef flat flagstone is little colonised by corals which mainly 
develop near the reef front. Acropora and Pocillopora are the 2 best 
represented genera. 

Molluscs 

The malacological fauna of Mataiva, with 222 species recorded (Table 
D), is very unequally distributed between the lagoon and the outer reef. 

On the outer reef, 169 species (156 gastropods and 13 bivalves) have 
been recorded. This fauna is fairly similar to those of other Tuamotu 
outer reefs. Nerita plicata, Nodolittorina leucosticta and Littorinea 



13 

coccinea occupy the upper zone, where the scarcity of Tectarius 
grand inatus is rather surprising. The reef flat flagstone harbours 
nearly 30 species of which a large majority belongs to a carnivorous 
epifauna (Vasum, Conus, etc.). A wide variety of forms can be collected, 
although none is abundant except Cypraea moneta . The fauna of the 
forereef is dominated by Turbo setosus , actively exploited by the atoll's 
inhabitants. Live Drupa morum and Cypraea caput-serpentis are abundant 
there as well as numerous shells coming from the outer slope, evidence of 
its richness. 

In the lagoon, 77 species have been collected: 55 gastropods and 22 
bivalves. The species distribution is very uneven: the richest fauna 
occurs near the hoa in the S of the atoll. Here, the more abundant 
species belong to the epifauna (Cypraeidae, Buccinidae): up to 20 Cyprea 
obvelata per m under coral blocks; on the sandy shoals, Cypraea moneta 
is the only macroraollusc found. One boring species, Lithophaga 
cinnamomina , inhabits the Porites colonies, at the rate of several dozen 
individuals per dm of the living surface. There is also a difference 
between the windward and leeward coasts of the same reef: Area 
ventricosa , Pinctada margaritif era and Tridacna maxima are more abundant 
on the leeward side. 

Moving away from the S of the atoll, there is a considerable decline 
in the malacological fauna. The sandy shoals are occupied by a few 
Cardium fragrum beds. A small oyster, Crassostrea cucculata , is 
particularly abundant on the branches of dead Acropora . The endofauna of 
bottom sediments is also unusual, consisting of many species unknown 
elsewhere in French Polynesia. 

Along the lagoon and raotu edges, the malacological fauna shows 
little diversity, only 3 species being found there. The absence of the 
Nasses, Mitres, Cones and Terebres usually found in this zone is 
surprising. 

On the littoral fringe of the motu, 3 species appear in succession: 
Littorina coccinea , Nerita plicata and Clypeomorus brevis , whilst the 
Cerithidae and Cypraeidae usually present in the upper levels are absent. 
In the same way, the hoa contain an impoverished homogeneous mollusc 
fauna (8 species) which is characteristic of semifunctional hoa 
(functioning intermittenty) . 

Mataiva's malacological fauna, which includes about 1/5 of the 
species recorded in French Polynesia (Richard, 1982), appears to be, on 
the whole, fairly rich. However, the lagoon fauna is, on the whole, poor, 
especially away from the zones under oceanic influence. Only a few 
species ( Lithophaga cinnamomina , Crassostrea cucculata ) are really very 
abundant. However, this malacological fauna is unusual because of the 
uncommon species found in the sediments and on the reef flats and the 
wide heterogeneity of the populations. 

Crustacean fauna 

The crustacean fauna of Mataiva atoll includes about 100 species 
(Monteforte, 1984). Most studies have been carried out on crabs, 



14 

especially coral-associated species, and on mud-shrimps, which are very 
numerous on the sandy shoals. 

Crustacea inhabit all environments in Mataiva. Land crabs ( Cardisoma 
carnif ex ) are plentiful in the coconut groves, as are hermit crabs 
Coenobita perlatus . The coconut crab Birgus latro , on the contrary, is 
rare (probably because of over-collecting) . 

The sandy environments of the lagoon are densely occupied by the 
mud-shrimps Callichirus armatus living in permanent burrows. Densities 
up to 3 mud-shrimps per square metre were observed on the muddy bottoms 
of the basins; they are lower in the shallow waters (0.5 to 1 ind./m ). 
For the whole lagoon, the estimated population of C_. armatus ranges from 
2.38 to 4.76 10 individuals, i.e. a biomass of 285 to 571 tons wet 
weight (100 to 200 kg per hectare). Callichirus armatus is a great 
sediment reworker, feeding on the mud which falls into its gallery, and 
generating considerable disturbance. It has been estimated that the 
upper centimetre of mataiva's bottom sediment passes through the 
mud-shrimps' burrows 4 to 9 times per year. Living also in a burrow, the 
storaatopod Lysiosquilla maculata ("varo") is common and much prized 
because of its tasty flesh. A crab Calappa hepatica inhabits these sandy 
surfaces, being well adapted to this biotope. The hard substrates in the 
lagoon, and especially Porites microatolls, shelter a reduced fauna, 
dominated by 2 Xanthidae, Chlorodiella nigra and Phymodius ungulatus , and 
1 Portunidae, Thalamita admete , more numerous near the hoa. 

The transition zones (hoa and pass) and the outer reef flats harbour 
a much more abundant and diversified population than does the lagoon. 
The hoa and the pass contain many species of the outer reef flat, which 
occasionally enter the lagoon: Chlorodiell cytherea , Liomera be! la , 
Pilodius pugil . One also finds many juveniles of outer reef species: 
Etisus laevimanus , Thalamita and Leptodius . The Grapsidae and Paguridae 
are dominant in the slightly sumserged zones. 

On the outer reef flats, the fauna is equally abundant and 
diversified, particularly on the reef front (Table E). Many species, 
such as the crayfish, Palunirus penicillatus , climb the outer slope at 
night and are found on the crest. The exposure of the reef flat modifies 
the crustacean fauna and its distribution; thus, Plagusia speciosa is 
abundant on the whole exposed reef flat but less commonly found on 
sheltered reefs and there limited to the reef front. 

On these outer reefs, the presence of living corals induces the 
existence of a well developed symbiotic fauna. This constitutes up to 
80% of the individuals associated with a living coral. The Xanthidae 
among the Brachyoura and the Alpheidae among the Natantia, are most 
abundant. In particular, crabs of the genus Trapezia (T_. speciosa , T_. 
bella and T_. f ormosa ) restricted to the coral Pocillopora , are 
particularly numerous in the frontal zones. 

Qther marine invertebrates 

At present, only the major invertebrate groups have been studied. 
However, some interesting observations on other groups may be cited. 



15 

.The sponges are abundant in the lagoon under the empty Tridacna 
shells or covering the dead Acropora branches. Many species with 
brightish colours are present, but the colonies are usually small, except 
for one black species, sometimes over 20 cm high, which is found in the 
lagoon and on the outer reef slope. 

The echinoderms include few species but some of them are very 
abundant. Such is the case, in the lagoon, of the black sea-cucumber, 
Halodeima atra , with densities reaching 1 or 2 individuals per m . This 
species is known to prefer more or less confined environments (Salvat, 
1975; Salvat et al.,1979). Another sea cucumber, Gucumaria sp., is often 
found near the hoa under empty Tridacna shells. 

On the outer reef, one may find two other Holothurians, the thin, 
long Synaptes near the beach, and the rounded, white spotted Bohadschia 
argus on the reef flat flagstone. But urchins are more abundant there. 
Echinometra mathaei is present in the reef flat cavities; near the reef 
front, the pencil urchin Heterocentrotus mamillatus is often found at the 
base of the spurs, while the helmet urchin, Colobocentrotus pedifer , 
whose morphology is well adapted to resist wave action, is more abundant 
on the upper part of the spurs, especially the relict algal ridge of the 
swell-exposed southern reef fronts. 

Ascidians form very discrete colonies and are not well known in 
French Polynesia. However, symbiotic ascidians, associated with green 
algae ( Prochloron ) , are present in Mataiva lagoon on the dead Acropora 
branches, and at the base of the spurs on the outer reef. 

Fishes 

The ichthyological fauna of Mataiva lagoon is poor, not only in 
number of species, but also in number of individuals. Of the 115 
species recorded (Table F), only 10 seem well established in the lagoon 
and are in evidence at practically all the stations. These are usually 
small individuals belonging to the Gobiidae ( Amblygobius phalaena , A. 
nocturnus ), the Pomacentridae ( Chromis coerulea ) , the Chaetodontidae 
( Chaetodon auriga , C. ephippium ) , or juveniles of Mullidae and Scaridae 
( Scarus sordidus , Scarus sp.). Some species, common to the lagoons of 
more or less closed atolls, are not found in Mataiva: Arothron hispidus , 
Chromis dimidiatus . 

Near the hoa and the pass, the number of species reported increases 
considerably (40-52 species). Generally, there is much heterogeneity in 
the distribution and abundance of the fish populations of Mataiva lagoon. 
More detailed studies (Bell and Galzin 1984, Galzin 1985), carried out in 
1981 and 1983, have demonstrated the close relation existing between the 
abundance and diversity of the fish fauna (total number of species, 
number of species 250 m~ , number of individuals 250 m ) and the live 
coral cover. Changes in live coral cover, estimated to be as small as 
to < 2% and < 2 to 2 to < 5%, produced significant increases in the total 
number of species and the number of individuals 250 m . On the other 
hand, the reef complexity, which is the same for dead and living coral 
colonies, is without any influence on fish populations. 



16 

In the pass and on the outer reef flats, the ichthyological fauna is 
much richer and made up of numerous species from the outside but which do 
not pass into the lagoon, such as triggler fishes ( Balistoides undulatus , 
Pseudobalistes f alvomarginatus , jacks ( Caranx trifasciatus ) ,emperors 
( Lethrinus mahsena ) , goatfishes ( Mugil angeli , _M. vaigiensis ) . 

On the outer slopes, the fish population is dominated by the 
Acanthuridae family, mainly Naso and Acanthurus . Pomacentridae, 
Serranidae, Lutjanidae and Chaetodontidae are equally abundant. The 
specific richness is at its maximum between 10 and 20 m (70 species), 
whereas the maximum abundance is found between 3 and 10 m. The effect of 
the 1983 cyclones on the outer reefs, causing massive coral destruction, 
has brought about a slight decrease of the herbivorous populations, but, 
above all, a redistribution of certain species and a much higher density 
in the upper levels of 3 to 10 ra. 



CONCLUSIONS 

Mataiva atoll has a singular morphology whose major characteristic is 
the partitioning of the lagoon into numerous pools. This is due to its 
peculiar geological history, during which several periods of uplift and 
subsidence occurred. During the periods of emergence, erosion processes 
resulted in the formation of a karstic relief, in the cavities of which 
phosphate deposits accumulated. 

The present morphology, moulded onto the former one, has, as a 
consequence, created particular hydrological conditions in the lagoon: 
very high turbidity, considerable variations in water level, high 
nutrient concentrations. 

The biological communities of the lagoon show the characteristics of 
a closed environment: few species are present, but some are very 
abundant. Mataiva stands out among atolls because of its high level of 
primary production, abundant zooplankton and a fairly poor, but very 
uneven distributed, benthic macrofauna. This latter, subject to strong 
variations in hydrological conditions, can suffer enormous mortality 
levels, affecting especially the corals. However, the evolution observed 
since 1981 seems to indicate that this is an accidental phenomenon and 
that the lagoon's biological communities retain the capability to survive 
and grow under difficult conditions. 

Future research on Mataiva atoll must take into account the wide 
range of variation in the distribution and abundance of its lagoonal 
populations. This fact seems to be closely related to the hydrological 
environment and its long-term variations. Such research will mainly 
concern the hydrology, the primary producers and a quantitative 
evaluation of the benthic and ichthyological fauna. 

Although Mataiva seems to be a very special atoll whose chara- 
cteristics cannot be used as a model for the other Tuamotu atolls, it is 
a very interesting experimental field for some ecological studies, e.g. 
the relationship between live coral cover and reef fish populations. A 
fish survey, similar to those made in 1981 and 1983, is already planned 
for mid-1985, to follow the changes in the fish communities, as the 



16 (Erratum, ARB 286) 

In the pass and on the outer reef flats, the ichthyological fauna is 
much richer and made up of numerous species from the outside but which do 
not pass into the lagoon, such as triggler fishes ( Balistoides undulatus, 
Pseudobalistes falvomarginatus , jacks ( Caranx trifasciatus ) , emperors 
( Lethrinus mahsena) , goatfishes ( Mugil angeli , M. vaigiensis ). 

On the outer slopes, the fish population is dominated by the 
Acanthuridae family, mainly Naso and Acanthurus . Pomacentridae, 
Serranidae, Lutjanidae and Chaetodontidae are equally abundant. The 
specific richness is at its maximum between 10 and 20 m (70 species), 
whereas the maximum abundance is found between 3 and 10 m. The effect of 
the 1983 cyclones on the outer reefs, causing massive coral destruction, 
has brought about a slight decrease of the herbivorous populations, but, 
above all, a redistribution of certain species and a much higher density 
in the upper levels of 3 to 10 m. 



CONCLUSIONS 

Mataiva atoll has a singular morphology whose major characteristic is 
the partitioning of the lagoon into numerous pools. This is due to its 
peculiar geological history, during which several periods of uplift and 
subsidence occurred. During the periods of emergence, erosion processes 
resulted in the formation of a karstic relief, in the cavities of which 
phosphate deposits accumulated. 

The present morphology, moulded onto the former one, has, as a 
consequence, created particular hydrological conditions in the lagoon: 
very high turbidity, considerable variations in water level, high 
nutrient concentrations. 

The biological communities of the lagoon show the characteristics of 
a closed environment: few species are present, but some are very 
abundant. Mataiva stands out among atolls because of its high level of 
primary production, abundant zooplankton and a fairly poor, but very 
uneven distributed, benthic raacrofauna. This latter, subject to strong 
variations in hydrological conditions, can suffer enormous mortality 
levels, affecting especially the corals. However, the evolution observed 
since 1981 seems to indicate that this is an accidental phenomenon and 
that the lagoon's biological communities retain the capability to survive 
and grow under difficult conditions. 

Future research on Mataiva atoll must take into account the wide 
range of variation in the distribution and abundance of its lagoonal 
populations. This fact seems to be closely related to the hydrological 
environment and its long-term variations. Such research will mainly 
concern the hydrology, the primary producers and a quantitative 
evaluation of the benthic and ichthyological fauna. 

Although Mataiva seems to be a very special atoll whose chara- 
cteristics cannot be used as a model for the other Tuamotu atolls, it is 
a very interesting experimental field for some ecological studies, e.g. 
the relationship between live coral cover and reef fish populations. A 
fish survey, similar to those made in 1981 and 1983, is already planned 
for mid-1985, to follow the changes in the fish communities, as the 
corals recover from almost complete destruction in 1980. 



17 



REFERENCES 

BELL, J. D., GALZIN, R. , 1984 - Influence of live coral cover on coral- 
reef fish communities. Mar. Ecol. Prog. Ser., 15 : 265-2 74. 

BUIGUES, D., 1983 - Sedimentation et diagenese des formations carbona- 

tees de l 1 atoll de Mururoa (Polynesie Francaise). These, 
Univ. Paris X, 2 vol., 309 p. 

BOURROUILH - LE JAN, F., 1980 - Phosphates, sols bauxitiques et karsts 

dolomitiques du Centre et du Sud-Ouest Pacifique. Doc. BRGM, 
24 : 113-128. 

CHEVALIER, J. P., DENIZOT, M. , 1979 - Les organismes constructeurs du 
lagon de Takapoto. J. Soc. Oceanistes, 35(65) : 31-34. 

DELESALLE, B., 1982 - Un atoll et ses problemes : Mataiva et ses phos- 
phates. Oceanis, 8(4) : 329-337. 

DELESALLE, B., BAGNIS, R. , BENNETT, J., BELL, J., DENIZOT, M. , GALZIN, 
R. , MONTAGGIONI, L. , PAYRI, C, RENON, J. P., RICARD, M., VER- 
GONZANNE, G. , 1983 - Biology, hydrology and ge omorphology of 
the atoll of Mataiva (Tuamotu Archipelago, French Polynesia). 
15 th Pacif. Sci. Ass. Cong., Dunedin, 1983, Abst. : 59. 

GALZIN, R. , 1985 - Ecologie des poissons recifaux de Polynesie 
Francaise. These, Univ. Montpellier, 195 p. 

HARMELIN-VIVIEN, M. , LABOUTE, P., 1983 - Preliminary data on under- 
water effects of cyclones on the outer slopes of Tikehau 
island (Tuamotu, French Polynesia) and its fish fauna. 3 r 
Intern. Soc. Reef Stud. Symp. , Nice, 1983, Abst. : 26. 

JARRIGE, F., GUEREDRAT, J. A., RECY, M., RAVAULT, P., 1978 - Etude de 
l 1 atoll de Mataiva. Rapport interne, 62 p. 

JOHNSON, M.W., 1954 - Plankton of the northtern Marshall Islands. U.S. 
Geol. Surv. Prof. Papers, 260 : 301-314. 

MICHEL, A., 1969 - Plancton du lagon et des abords exterieurs de 
l 1 atoll de Mururoa. Cah. Pacif. 13 : 81-132. 

MICHEL, A., COLIN, C. , DESROSIERES, R. , OUDOT, C. , 19 71 - Observations 

sur l'hydrologie et le plancton des abords et de la zone des 
passes de l 1 atoll de Rangiroa (Archipel des Tuamotu, Pacifi- 
que central). Cah. O.R.S.T.O.H. , ser. Oceanogr.,9 : 375-402. 

MONTAGGIONI, L.F., 1985 - Makatea island, Tuamotu archipelago. In : 

B. DELESALLE, R. GALZIN & B.SALVAT (Eds). 5 th International 
Coral Reef Congress, Tahiti 27 May - 1 June. Vol.1 :"French 
Polynesian Coral reefs": 103-158. 

MONTAGGIONI, L.F., PIRAZZOLI, P. A., 1984 - The significance of exposed 

coral conglomerates from French Polynesia (Pacific Ocean) as 
indicators of recent relative sea-level changes. Coral Reefs, 
3 : 29-42 

MONTEFORTE, M., 1984 - Etude des peuplements de Crustaces Decapodes 
Reptantia et Stomatopodes de Polynesie Francaise. These, 
Paris VI, 148 p. 

PIRAZZOLI, P. A., MONTAGGIONI, L. F., Geological factors controlling 

reef evolution in the north-western Tuamotu region. 
Quaternary Research, in press. 

RENON, J. P., 1977 - Zooplancton du lagon de Takapoto (Polynesie 
Francaise). Ann. Inst. Oceanogr. 53 : 217-236. 



18 



RICHARD, 

ROUGERIE 

ROUGERIE 
SALVAT, 

SALVAT, 

SOURNIA, 
VAUGELAS 



G., 1982 - Mollusques lagunaires et recifaux de Polynesie 
Francaise : inventaire faunistique, bionomie, bilan quanti- 
tatif, croissance, production. These, Paris VI, 313 p. 

F., 1983 - Nouvelles donnees sur le fonctionnement interne 
des lagons d' atoll. C. R. Acad. Sc. Paris, Serie II, t. 297 : 
909-912. 

F., RICARD, M., MAZAURY, E. , 1982 - Le lagon de l 1 atoll de 
Mururoa. Rapport C.E.A., R. 5236 : 92 p. 

B., 1975 - Qualitative and quantitative study of Halodeima 
atra , (Echinodermata, Holothuridea) , in the lagoons and reefs 
of French Polynesia. Proc. 13 Pacif. Sci. Cong., Vancouver, 
1975. Abstr. 1 : 132. 

, RICHARD, G., CHEVALIER, J. P., 
J., 1979 - Consequences ecologi- 
d'extraction de sable corallien 
(tie de la Societe, Polynesie 
1(1): 83-126. 

A., RICARD, M. , 1976 - Donnees sur l'hydrologie et la produc- 
tivity du lagon d'un atoll ferme (Takapoto, lies Tuamotu) . 
Vie et Milieu 26(2), Serie B : 243-279. 

J. de, 1983 - First record of the Callianassa (Crustacea, 
Thalassinidea) Callichirus armatus Milne-Edward 1870, in 

Int. 



B., VERGONZANNE, G., GALZIN, R 
RICARD, M., RENAUD-MORNANT, 
ques des activites d'une zone 
dans le lagon de Moorea 
Francaise). Cah. Indo-Pacif., 



»rd 



Polynesian Islands (Tahiti, Moorea and Mataiva) 
Soc. reef Stud. Symp., Nice, Abstr. :23. 
VAUGELAS J. de, DELESALLE, B., MONIER, C. , - Aspects of the biology of 

Cal lichirus armatus , A. Milne Edwards 1870 (Crustacea, Thalas- 
sinidea) from French Polynesia. Crustaceans, in press. 



19 



Table A : Distribution of the phy toplanktonic species in the different 
sectors of the lagoon. 





East 


North 


Pass 


Hoa 


Center 


DIATOMOPHYCEAE 










Achnanthes sp. 




X 






Actinoptychus undulatus 


X 




X 


X 


Amphiprora alata 


X 


X 


X 




Amphora bigibba 


X 


X 




X 


Amphora exsecta 


X X 


X 


X 




Amphora obtusa 


X 








Amphora ostrearia 


X 


X 






Amphora robusta 




X 






Amphora sp. 


X 


X 


X 


X 


Asterionella kariana 






X 




Asterolampra marylandica 








X 


Biddulphia sp. 








X 


Caloneis liber 


X 








Caloneis sp. 


X 


X 






Campylodiscus innominatus 


X 


X 


X 




Chaetoceros sp. 








X 


Climacosphenia moniligera 


X 




X 


X 


Cocconeis sp. 


X 






X 


C. placentula var.euglypta 


X 








Coscinodiscus cf eccentricus 


X 




X 


X 


Coscinodiscus radiatus 






X 




Cyclotella menneghinianna 


X 








Diploneis bombus 




X 






Hemidiscus cuneiformis 


X 




X 




Grammatophora marina 




X 






Gyrosigma corallinum 


X 








Gyrosigma sp. 


X 


X 






Lycmophora Ehrenbergii 


X 


X 


X 




Mastogloia affirmata 


X 








Mastogloia binotata 


X 


X 


X 


X 


Mastogloia corsicana 


X 


X 






Mastogloia decussata 


X 








Mastogloia erythrea 


X 


X 






Mastogloia fimbriata 


X 


X 




X 


Mastogloia horvathiana 


X X 


X 






Mastogloia occulata 


X 






X 


Mastogloia ovata 






X 




Mastogloia ovulum 


X X 


X 






Mastogloia cf pseudoparadoxa 


X 








Mastogloia splendida 


X 








Mastogloia cf tenuis 


X 








Mastogloia sp. 




X 






Navicula exigua 






X 




Navicula granulata 


X 








Navicula longa 




X 






Navicula lyra 


X 


X 







20 





East 


North 


Pass 


Hoa 


Center 


Navicula cf menisculus 




X 


X 






Navicula perobesa 


X 




X 


X 




Navicula sp. 






X 


X 




Nitzschia acuta 


X 


X 




X 




Nitzschia closterium 




X 


X 


X 


X 


Nitzschia distans 




X 








Nitzschia longissima 




X 


X 


X 


X 


N. punctata var coarctata 






X 




X 


Nitzschia sigma 






X 






Nitzschia seriata 








X 




Nitzschia ventricosa 




X 








Nitzschia sp. 




X 






X 


Plagiogramma atomus 








X 




Pleurosigma sp. 




X 




X 




Podocystis spathulata 


X 


X 


X 


X 


X 


Rhabdonema adriaticum 




X 








Rhizosolenia sp. 






X 


X 




Stauroneis salina 


X 




X 






Surirella fastuosa 


X 


X 




X 




Surirella reniformis 


X 


X 








Surirella sp. 


X 










Synedra ulna 






X 


X 


X 


Synedra undulata 


X 


X 






X 


Thalassiosira sp. 










X 


Trachyneis sp. 








X 




Triceratium shadboltianum 








X 




Tropidoneis lepidoptera 


X 


X 


X 


X 




DINOPHYCEAE 












Ceratium pentagonum 








X 




Dinophysis sp. 










X 


Exuviella sp. 






X 


X 




Gonyaulax spinifera 




X 






X 


Gymnodinium splendens 










X 


Gymnodinium sp. 


X 


X 


X 


X 


X 


Oxytoxum sp. 




X 


X 






Orni thocercus quadratus 








X 




Oxytoxum sp. 










X 


Prorocentrum sp. 






X 






Protoperidinium sp. 




X 


X 






CHLOROPHYCEAE 












Carteria sp. 


X 


X 


X 


X 




Chlamydomonas sp. 


X 


X 








Nephroselmis sp. 




X 


X 






Pyramimonas sp. 


X 


X 


X 


X 


X 


CYANOPHYCEAE 












Oscillatoria sp. 


X 


X 


X 






Pseudanabaena sp. 






X 


X 




Spirulina sp. 




X 


X 


X 




CRYPTOPHYCEAE 


X 


X 






X 


COCCOLITHOPHORACEAE 


X 






X 


X 



21 



Table B: Distribution of the main algal species in the lagoon and the 
outer reef. 





Outer 




Lagoon 






Reef 


Hoa 


Pass | East 


| North | Center 


CYANOPHYCEAE 










Calothrix sp. 




X 


X X 


X 


Hassalia byssoides 


X 


X 






Lyngbia majuscula 






X 




CHRYSOPHYCEAE 










Chrysonephros sp. 




X 


X 


X 


PHEOPHYCEAE 










Dictyota sp. 






X 




Lobophora variegata 


X 




X 




Sphacelaria furcigera 




X 




X 


Sph. tribuloides 




X 






Turbinaria ornata 


X 








CHLOROPHYCEAE 










Acetabularia moebii 






X 




Avrainvillea lacerata 


X 


X 


X X 


X X 


Boodlea sp. 




X 


X 




Bryopsis sp. 




X 


X 




Caulerpa peltata 




X 


X 


X 


Caulerpa racemosa 


X 




X 




Caulerpa serrulata 


X 


X 


X 


X X 


Caulerpa sp. 


X 




X 


X 


Codium adherens 




X 






Dictyosphaeria sp. 


X 


X 


X X 


X 


Enteromorpha sp. 




X 


X 




Halimeda discoidea 


X 








Halimeda opuntia 


X 


X 


X X 


X X 


Microdictyon sp. 


X 


X 






Neomeris sp. 


X 








Struvea elegans 






X 




Trichosolen sp. 




X 






RHODOPHYCEAE 










Centroceras clavulatum 




X 


X 




Ceramium tenerrimum 




X 


X 




Ceramium sp. 




X 


X 




Chevaliericrusta sp. 


X 








Gelidium crinale 


X 


X 


X 


X 


Gelidium pusillum 


X 


X 


X 




Herposiphonia secunda 




X 


X X 


X 


Hypnea sp. 




X 


X 




Jania sp. 


X 


X 


X 




Liagora decussata 


X 








Porolithon onkodes 


X 








Porolithon craspedium 


X 








Polys iphonia sp. 




X 


X 


X 


Pterocladia media 


X 


X 




X 



PSAMMOCORA CONTIGUA 
PSAMMOCORA SUPERFICIALIS 
POCILLOPORA DAMICORNIS 
POCILLOPORA EYDOUXI 
POCILLOPORA VERRUCOSA 
ASTREOPORA MYRIOPHTHALMA 
MONTIPORA AEQUITUBERCULATA 
MONTIPORA COMPOSITA 
MONTIPORA EDWARDS I 



OUTER 
REEF 



PASS 



LAGOON 

NORTH . SOUTH CENTER . EAST 

(J) (B) (L) (G) (C) 



MONTIPORA INFORMS 


+ 




MONTIPORA TUBERCULOSA 


+ 


+ 


MONTIPORA TURGESCENS 






MONTIPORA VERRUCOSA 




+ 


ACROPORA DANAI 


+ 




ACROPORA HUMILIS 


+ 


♦ 


ACROPORA LATISTELLA 




+ 


ACROPORA ROBUSTA 


+ 




ACROPORA TORTUOSA 






ACROPORA VALIDA 




+ 


PAVONA MINUTA 


+ 




PACHYSERIS SPECIOSA 


+ 




FUNGIA FUNGITES 


+ 




PORITES (SYNARAEA) CONVEXA 




+ 


PORITES (NAPOPORA) IRREGULARIS 


+ 




PORITES LICHEN 


+ 




PORITES LOBATA 


+ 


+ + 


PORITES CF SOLIDA 




+ 


FAVIA PALLIDA 




+ 


FAVIA ROTUMANA 


+ 




FAVIA STELLIGERA 


+ 




LEPTASTREA PURPUREA 


+ 


+ + 


LEPTASTREA TRANSVERSA 






CYPHASTREA SERAILA 




+ 


PLATYGYRA DAEDALEA 




+ 


MONTASTREA CURTA 


+ 




ACANTHASTREA ECHINATA 




+ 


LOBOPHYLLIA CORYMBOSA 


+ 




Dead Colonies 







Table C : Distribution of the main species of Sclerac tinians on the 
outer reef and in the lagoon. 



23 

Table D: List of Molluscs catalogued in the lagoon and on the outer reef 

CLASS GASTROPODA 

SUB-CLASS PROSOBRANCHIA 

Order Archaeogastropoda 

HALIOTIDAE 

Haliotis pulcherrima Gmelin, 1791 
PATELLIDAE 

Patella flexuosa Quoy & Gaimard, 1834 
STOMATELLIDAE 

Stomatella sanguinea (Adams, 1850) 

Stomatella varia (Adams, 1850) 
TURBINIDAE 

Turbo argyrostomus Linne, 1758 

Turbo petholatus Linne, 1758 

Turbo setosus Gmelin, 1791 
NERITIDAE 

Clithon chlorostoma (Broderip, 1832) 

Nerita plicata Linne, 1758 

Puperita reticulata (Sowerby, 1832) 
Order Mesogastropoda 
NEOCYCLOTIDAE 

Amphicyclotus sp. 
LITTORINIDAE 

Littorinea coccinea (Gmelin, 1791) 

Nodilittorina leucosticta (Reeve, 1857) 

Tectarius grandinatus (Gmelin, 1791) 
TRUNCATELLIDAE 

Truncatella sp. 
VERMETIDAE 

Dendropoma maximum (Sowerby, 1825) 

Serpulorbis sp. 
PLANAXIDAE 

Planaxis lineatus (Da Costa, 1776) 
MODULIDAE 

Modulus tectum (Gmelin, 1791) 
CERITHIDAE 

Bittium cf. glareosum (Gould, 1861) 

Bittium zebrum (Kiener, 1841) 

Cerithium atromarginatum Dautzenberg & Bouge, 1933 

Cerithium columna Sowerby, 1834 

Cerithium mu tat urn Sowerby, 1834 

Cerithium nesioticum Pilsbry & Vanatta, 1906 

Cerithium rostratum Sowerby, 1855 

Cerithium salebrosum Sowerby, 1855 

Clypeomorus brevis Quoy & Gaimard, 1834 

Clypeomorus moniliferus (Kiener, 1841) 

Rhinoclavis diadema Houbrick, 1878 

Rhinoclavis sinensis (Gmelin, 1791) 



24 



EULIMIDAE 

Eulima sp. 
STROMBIDAE 

Strombus dentatus Linne, 1758 

Strombus gibberulus Linne, 1758 

Strombus maculatus Sowerby, 1842 

Strombus mutabilis Swainson, 1821 
CALYPTRAE IDAE 

Cheila equestris (Linne, 1758) 
TRIVIIDAE 

Trivia sp. 
CYPRAEIDAE 

Cypraea caputserpentis Linne, 1758 

Cypraea carneola var. propinqua Garrett, 1879 

Cypraea cumingii Sowerby, 1832 

Cypraea depressa Gray, 1824 

Cypraea dillwyni Schilder, 1922 

Cypraea erosa Linne, 1758 

Cypraea fimbriata Gmelin, 1791 

Cypraea goodalli Sowerby, 1832 

Cypraea helvola Linne, 1758 

Cypraea irrorata Gray, 1828 

Cypraea isabella Linne, 1758 

Cypraea leviathan Schilder & Schilder, 1937 

Cypraea maculifera (Schilder, 1932) 

Cypraea margarita Dillwyn, 1817 

Cypraea minoridens Melvill, 1901 

Cypraea mone ta Linne, 1758 

Cypraea nucleus Linne, 1758 

Cypraea obvelata Lamarck, 1810 

Cypraea poraria Linne, 1758 

Cypraea schilderorum (iredale, 1939) 

Cypraea scurra Gmelin, 1791 

Cypraea serrul ifera Schilder & Schilder, 1938 

Cypraea subteres Weinkauff, 1880 

Cypraea talpa Linne, 1758 

Cypraea tigris Linne, 1758 

Cypraea ventriculus Lamarck, 1810 
NAT IC IDAE 

Natica galteriana Recluz, 1844 

Polinices melanostoma (Gmelin, 1791) 
CYMATIDAE 

Cymatium gemmatum (Reeve, 1844) 

Cymatium hepaticum (Rbding, 1798) 

Cymatium muricinum (Rdding, 1798) 

Cymatium nicobaricum (Roding, 1798) 

Cymatium rubeculum (Linne, 1758) 

Distortio anus (Linne, 1767) 

Gyrineum roseum (Reeve, 1844) 



25 



BURSIDAE 

Bursa bufonia (Gmelin, 17 91) 
Bursa granular is (Roding, 1798) 

COLUBRARIIDAE 

Colubraria nitidula (Sowerby, 1833) 
Colubraria tortuosa (Reeve, 1844) 
Colubraria sp. 
Order Neogastropoda 

MURICIDAE 

Drupa clathrata (Lamarck, 1816) 

Drupa grossularia Roding, 1798 

Drupa morum (Roding, 1798) 

Drupa ricinus (Linne, 1758) 

Drupa speciosa (Dunker, 1867) 

Drupel la cornus (Roding, 1798) 

Drupella fenestrata (Blainville, 1832) 

Drupel la ochrostoma (Balinville, 1832) 

Homalocantha martinetana (Roding, 1798) 

Maculotriton serriale (Deshayes, 1834) 

Mancinella tuberosa (Roding, 1798) 

Morula granulata (Duclos, 1832) 

Morula margariticola (Broderip, 1832) 

Morula uva (Roding, 179 8) 

Nassa francolinus (Bruguiere, 1789) 

Pterinotus loebbeckei (Kobelt, 1879) 

Thais aculeatus (Deshayes & Milne-Edwards, 1844) 

Thais armigera (Link, 1807) 

CORALLIOPHILIDAE 

Coralliophila cf. porphyroleuca (Crosse 1870) 
Coralliophila violacea (Kiener, 1836) 
Leptochoncus lamarcki Deshayes, 1863 
Quoyula madreporarum (Sowerby, 1832) 

BUCCINIDAE 

Cantharus fumosus (Dillwyn, 1817) 

Cantharus spica (Melvill & Standen, 1895) 

Cantharus undosus (Linne, 175 8) 

Engina incarnata (Deshayes, 1834) 

Engina sp. 

Pisania decollata (Sowerby, 1833) 

Pisania ignea (Gmelin, 1791) 

Pisania iostoma (Gray, 1834) 

Pisania truncata (Hinds, 1844) 

COLLUMBELLIDAE 

Mitrella sp . 

Pyrene flava (Bruguiere, 1789) 

Pyrene scripta (Lamarck, 1822) 

NASSARIIDAE 

Nassarius gaudiosus (Hinds, 1844) 
Nassarius cf. pauperus (Gould, 1850) 



26 



FASCIOLARIIDAE 

Latirus sanguifluus (Reeve, 1847) 

Latirus sp. 

Peristernia chlorostoma (Sowerby, 1825) 

Peristernia nassatula (Lamarck, 1822) 

Peristernia sp. 
VASIDAE 

Vasum ceramicum (Linne, 1758) 
HARP IDA E 

Harpa gracilis Broderip & Sowerby, 1829 
MITRIDAE 

Imbricaria punctata (Swainson, 1821) 

Mitra assimilis Pease, 1868 

Mitra coffea Schubert & Wagner, 1829 

Mitra columbellif ormis Kiener, 1838 

Mitra cucumerina Lamarck, 1811 

Mitra fastigium Reeve, 1845 

Mitra ferruginea Lamarck, 1811 

Mitra litterata Lamarck, 1811 

Mitra paupercula (Linne, 1758) 

Mitra pell isserpentis Reeve, 1844 

Mitra stictica (Link, 180 7) 
COSTELLARIIDAE 

Thala mirifica (Reeve, 1845) 

Thala sp. 

Vexillum cadaverosum (Reeve, 1844) 

Vexillum crocatum (Lamarck, 1811) 

Vexillum cumin gi i (Reeve, 1844) 

Vexillum speciosum (Reeve, 1844) 
TURRIDAE 

Clavus formosus (Reeve, 1847) 

Daphnella sp. 

Lienardia rubida (Hinds, 1844) 

Lienardia cf roseotincta (Montrouzier , 1872) 

Xenuroturris cingulifera (Reeve, 1847) 
CON I DAE 

Conus auratinus Da Motta, 1982 

Conus auricomus Hwass j^ Bruguiere, 1792 

Conus catus Hwass Jji Bruguiere, 1792 

Conus chaldaeus (Rdding, 1798) 

Conus cylindraceus Broderip & Sowerby, 1830 

Conus distans Hwass _in Bruguiere, 1792 

Conus ebraeus (Linne, 175 8) 

Conus flavidus Lamarck, 1810 

Conus geographus Linne, 1758 

Conus legatus Lamarck, 1810 

Conus lividus Hwass _in_ Bruguiere, 1792 

Conus magnificus Reeve, 1843 

Conus miles Linne, 1758 

Conus miliaris Hwass _in Bruguiere, 1792 

Conus pertusus Hwass in Bruguiere, 1792 



27 



Conus pulicarius Hwass in Bruguiere, 1792 

Conus rattus Hwass in Bruguiere, 1792 

Conus retifer Menke, 1829 

Conus scabriusculus Dillwyn, 1817 

Conus sponsalis Hwass _in Bruguiere, 1792 

Conus tenuistriatus Sowerby, 1858 

Conus textilinus Kiener, 1845 

Conus tulipa Linne, 1758 

Conus vex ilium Gmelin, 1791 

TEREBRIDAE 

Terebra crenulata (Linne, 1758) 
Terebra guttata (Roding, 1798) 
Order Heterogastropoda 

ARCHITECTONIC IDAE 

Heliachus infundibulif ormis (Gmelin, 1791) 

EPITONIIDAE 

Epitonium sp. 

JAN THIN IDAE 

Janthina ianthina (Linne, 1758) 

TRIPHORIDAE 

Triphora sp. 
SUB-CLASS OPISTHOBRANCHIA 
Order Entomotaeniata 

PYRAMIDELL IDAE 

Pyramidella sp. 
Order Cephalaspidea 

ACTEONIDAE 

Pupa solidula (Linne, 1758) 

HYDATINIDAE 

Hydatina amplustre (Linne, 1758) 

ATYIDAE 

Atys sp. 
SUB-CLASS PULMONATA 

Order Basommatophora 

ELOBI IDAE 

Melampus sp. 
CLASS BIVALVIA 

Order Arcoida 

ARC IDAE 

Area imbricata Bruguiere, 1789 
Area ventricosa Lamarck, 1819 
Order Mytiloida 

MYTIL IDAE 

Lithophaga cinnamomina (Chemnitz, 1785) 
Modiolus auriculatus Krauss, 1848 

PINNIDAE 

Pinna muricata Linne, 1758 

PTERIIDAE 

Pinctada maculata (Gould, 1850) 
Pinctada margaritif era (Linne, 1758) 



28 



ISOGNOMONIDAE 

Isognomon sp. 

PECTINIDAE 

Chlamys inaequivalvi s (Sowerby, 1842) 
Chlamys sp. 

OSTREIDAE 

Crassostrea cucullata (Born, 1778) 
Order Hippuritoida 

CHAM I DAE 

Cham a iostoma Conrad, 183 7 
Chama pacifica Broderip, 1834 
Order Veneroida 

LUCINIDAE 

Anodontia edentula (Linne, 1758) 
Codakia puncata (Linne, 1758) 
Codakia divergens (Philippi, 1850) 

CARDIIDAE 

Corculum fragum (Linne, 1758) 

TRIDACNIDAE 

Tridacna maxima (Rbding, 1798) 

TELLINIDAE 

Arcopagia robusta (Hanley, 1844) 
Quidnipagus palatam Iredale, 1929 
Scutarcopagia scobinata (Linne, 1758) 
Tellina donac iformis Deshayes, 1854 
Tellina obliquaria Deshayes, 1854 

PSAMMOBI IDAE 

Asaphis violaceus (Forskal, 1775) 

TRAPEZIIDAE 

Trapezium oblongum (Linne, 1758) 

VENERIDAE 

Gafrarium pectinatum (Linne, 1758) 
Pi tar prora (Conrad, 183 7) 



a i — i 


s 


M <C 


S cn «: w p w -* 

W <I 1-1 D W O LO E-t OS 


O <[£lco tomOMWcn HWO o eh M J 


Hcohwco o 3 w s«i o p CO en co to -3 o E-< < <: co < 

<c to o <: w m m w eh o h joe: o to w o h h < q ^mk 

SOHMM> a. O 2 =5 O < 5 MO to&3PL,coto:DU <JO .JOH 


O X < c*5 H W to ,-1 w M o H OHQ n<HMa:oKM to 03 M — <N M CC p: 


PSIS SI 
S CONVE 
S MACUL 
ELLA BA 
ELLA CY 
ELLA LA 
RLATA 
SEBANA 
LUMNUS 
UMNUS G 
S SANGU 
LOIDES 
BELLA 
ODES AL 
NULIPES 
HIAS NO 
PUGIL 
SCABRI 
US CAYS 
TETRAO 
ENEUS 
ANICUL 
LAEVIM 
A PERLA 
US CRIN 
LONGITA 
TENUICR 
PSUS MI 
PSUS PL 
AFFINE 

SPECIO 
S ERYTH 
A DAKIN 
A PICTA 
THES SP 
THES SP 
S PENIC 
US HOLT 
CERATOP 


ATERGATO 

CARPILIU 

CARPILIU 

CHLORODI 

CHLORODI 

CHLORODI 

DAIRA PE 

ERIPHIA 

GLABROPI 

GLOBOPIL 

LEPTODIU 

LIOCARPI 

LIOMERA 

LIOXANTH 

LYDIA AN 

PARAXANT 

PILODIUS 

PILODIUS 

PSEUDOZI 

XANTHIAS 

ZOYMUS A 

ANICULUS 

CALCINUS 

COENOBIT 

GEOGRAPS 

GRAPSUS 

GRAPSUS 

PACHYGRA 

PACHYGRA 

PERCNON 

PLAGUSIA 

CHARYBDI 

THALAMIT 

THALAMIT 

PETROLIS 

PETROLIS 

PANULIRU 

PARRIBAC 

OCYPODE 


— w(n>TLnvDr^to^o — Nfnstnor^co^o — Ncn-^i/ivDrvajoio — Nn-riooi — co ct» 


c 


/ o ! 


CO 


ft .=. 




O 


// r 






o 


( S 


o 


CN ! 

CN CO 




/(/ " 










/ / n 


CO ^ 














CN 

CN C 


) o 


O 

CN T~ 

o[ 










•- CO 


00 
CO 


CO 
CO 

D 




\ u 


















L 






\i\ *- 




















I 
















v 


-1 1 














r 


















\ 


~| 










-1 






if r 










|- 








1! 
















I 














I ■ 






J 








* 1 


k r 


2 








If 












1 






en 








i— 1 


<N 








& 


o 






n 




\j 














v i y 




00 










f 














i 




























e> 








, : i 


r 


~| 








"1 


•o 

CN 




V , 
































m 






O 


r> 


» 


> in co 


00 




"— ■- CO CN CN — CN 


fSI 






S 


"1 




fr^ff! 






//"^ /■ 






Wrv" 


o 

CO 














% 


in iv 




CN CN 




!.< 
















O 


': L 




< '-■ 





u 

01 

•p 

3 
O 

-a 

CD 

u 

CD 



CD 



CD 



D 
U 

o 

a) 

x; 



M-l 

o 



to 

P. 
CD 



<u 

cd 
H 



30 

Table F: List of fishes catalogued in Mataiva Lagoon (L) and the nearby 
ocean (0) . 

CLASS CHONDRICHTHYES 

Order Carcharhiniformes 

CARCHARHINIDAE 

Carcharhinus melanopterus (Quoy et Gaimard ,1824) L 
Order Rajifonnes 
MYLIOBATIDAE 

Aetobatis narinari (Euphrasen,l 790) L 

CLASS OSTEICHTHYES 

Order Angui 1 1 if ormes 
HJRAENIDAE 

Echidna nebulosa (Ahl,1789) L 

Gymnothorax javanicus (Ble eker , 185 9) L 

Gymnothorax meleagris (Shaw et Nadder,1795) L 

Order Au lop if ormes 
SYNODONTIDAE 

Saurida gracilis (Quoy et Gaimard ,1824) L 

Order Gadi formes 
OPHIDIIDAE 

Dinematichthys sp. L 

Order Atherinif ormes 
HEMIRHAMPHIDAE 

Hyporhampus acutus (Giinther ,1871) L 

BELONIDAE 

Tylosurus crocodilus (Lesueur ,1821) L 

Order Berycyformes 
HOLOCENTRIDAE 

Myripristis kuntee Cuvier,1831 L 

Myripristis murdjan (Forsskal , 177 5) L 

Myripristis sp.417 L 

Neoniphon argenteus Bleeker,1849 L 

Neoniphon opercularis Valenciennes , 1831 L 

Neoniphon samara (ForsskSl , 1775) L 

Sargocentron caudimaculatum (Riippel 1 , 1826) 

Sargocentron microstoma Giinther ,1859 L 

Order Syngnathi formes 
FISTULARIIDAE 

Fistularia commersonii Ruppell,1838 L 

Order Scorpaeniformes 
SCORPAENIDAE 

Scorpaenodes guamensis (Quoy et Gaimard ,1824) L 

Order Perciformes 
SERRANIDAE 

Anthias pascalus (Jordan et Tanaka,1927) 

Anthias squamipinnis (Peters , 1855) 

Cephalopholis argus (Bloch et Schneider ,1801) L 

Cephalopholis urodelus (Bloch et Schne ider ,1801) 
Epinephelus merra Bloch, 1793 L 

Epinephelus microdon (Bleeker ,1856) L 

Gracila albomarginata (Fowler et Bean, 1930) 



31 

GRAMMISTIDAE 

Grammistes sexlineatus Thumberg ,1792 L 

PSEUDOGRAMMIDAE 

Pseudogramma polycantha (Bleeker ,1856) L 

APOGONIDAE 

Apogon exostigma (Jordan et Starcks , 1906) L 

Apogon kallopterus Bleeker, 1856 L 

Apogon novemfasciatus (Cuvier ,1828) L 

Apogon savayensis (Gvinther ,1871) L 

Cheilodipterus lineatus (Lacepede, 1801) L 

Cheilodipterus macrodon (Lacepede, 1802) L 

Cheilodipterus quinquelineatus (Cuvier ,1828) L 

Fowleria aurita Valenciennes, 1831 L 

Fowleria marmoratus Alleyne et MacLeay,1876 L 

Fowleria sp.297 L 

Genus sp.289(juv.) L 

Genus sp.292 (juv.) L 

CARANGIDAE 

Caranx ignobilis (Forsskal ,1775) 

Caranx melampygus (Cuvier ,1833) L 

Gnathanodon speciosus (Forsskal ,1775) L 

LUTJANIDAE 

Aphareus furca (Lacepede, 1802) 

Lutjanus bohar (Forsskal ,1775) 

Lutjanus fulvus (Bloch et Schneider ,1801) L 

Lutjanus gibbus (Forsskal ,1775) L 

LETHRINIDAE 

Lethrinus xanthochilinus (Klunzinger ,1870) L 

Monotaxis grandoculis (Forsskal ,1775) L 

MULLIDAE 

Mulloides f lavolineatus (Lacepede, 1801) L 

Mulloides vanicolensis (Valenciennes ,1831) L 

Parupeneus barber inus (Lacepede ,1802) L 

Pseudupeneus bifasciatus (Lacepede, 1802) L 

Pseudupeneus multifasciatus (Lacepede, 1801) L 

CHAETODONTIDAE 

Chaetodon auriga Forsskal ,1775 L 

Chaetodon bennetti Cuvier, 1831 L 

Chaetodon citrinellus Cuvier, 1831 L 

Chaetodon ephippium Cuvier, 1831 L 

Chaetodon kleinii Bloch, 1790 L 

Chaetodon lineolatus Cuvier, 1831 L 

Chaetodon lunula (Lacepede, 1803) L 

Chaetodon ornatiss imus Cuvier, 1831 L 

Chaetodon pelewensis Kner,1868 

Chaetodon quadr imaculatus Gray, 1831 L 

Chaetodon reticulatus Cuvier, 1831 

Chaetodon semeion Bleeker, 185 5 L 

Chaetodon trifascialis Quoy et Gaimard,1824 L 

Chaetodon trifasciatus Park, 179 7 L 

Chaetodon ulietensis Cuvier, 1831 L 



32 



Chaetodon unimaculatus Bloch,1787 L 

Chaetodon vagabundus Linne,1758 L 

Forcipiger flavissimus Jordan et MacGregor ,1898 

Forcipiger longirostris (Broussonnet , 1782) 

POMACANTHIDAE 

Centropyge flavissimus (Cuvier ,1831) L 

Centropyge loriculus (Giinther ,1860) 

POMACENTRIDAE 

Abudefduf sexfasciatus (Lacepede , 1801) L 

Abudefduf sordidus (Forsskal ,1775) L 

Amphiprion clarkii (Bennett ,1830) 

Chromis caerulea (Cuvier ,1830) L 

Chromis iomelas Jordan et Seale,1906 

Chromis vanderbilti (Fowler , 1941) 

Chromis sp.372 L 

Chrysiptera leucopoma (Lesson ,1830) L 

Dascyllus aruanus (Linne,1758) L 

Dascyllus trimaculatus (Ruppell ,1828) 

Dascyllus reticulatus (Richardson , 184 6) 

Plectroglyphidodon dickii (Lienard,1839) 
Plectroglyphidodon flaviventris Allen et Randal 1,1974 L 
Plectroglyphidodon johnstonianus Fowler et Ball, 1924 

Pomacentrus pavo (Bloch,178 7) L 

Pomacentrus coelestis Jordan et Starks,1901 L 

Stegastes albifasciatus (Schlegel et Muller,1839) L 

Stegastes aureus (Fowler , 192 7) 

Stegastes nigricans (Lacepede, 1803) L 

CIRRHITIDAE 

Paracirrhites arcatus (Jordan et Evermann , 1903) 

Paracirrhites forsteri (Schne ider ,1801) 

Paracirrhites sp.420 (juv.) 

SPHYRAENIDAE 

Sphyraena barracuda (Walbaum, 1 792) L 

LABRIDAE 

Bodianus anthioides (Bennett ,1830) 

Bodianus sp.151 L 

Cheilinus chlorourus (Bloch,1791) L 

Cheilinus oxycephalus Bleeker,185 3 L 

Cheilinus trilobatus Lacepede, 1801 L 

Cheilinus undulatus Ruppell, 1835 L 

Cheilinus unifasciatus Streets, 1811 L 

Cirrhilabrus sp.58 

Coris aygula (Lacepede, 1801) L 

Coris gaimard (Quoy et Gaimard,1824) L 

Cymolutes sp. L 

Epibulus insidiator (Pallas, 1 770) L 

Gomphosus varius Lacepede, 1801 L 

Halichoeres hortulanus Lac e'pede, 1801 L 

Halichoeres marginatus Ruppell, 1835 L 

Halichoeres ornatissimus (Garrett ,1889) 

Halichoeres trimaculatus (Quoy et Gaimard ,1834) L 



33 



Halichoeres sp.(juv.) L 

Hemigymnus fasciatus (Bloch,1792) 

Labroides dimidiatus (Valenciennes, 1839) L 

Pseudocheilinus octo taenia Jenkins, 189 9 L 

Pseudocheilinus tetrataenia Schultz,1946 L 

Stethojulis bandanensis (Ble eker ,1851) L 

Thalassoma amblycephalum Bleeker,1856 L 

Thalassemia hardwicke (Bennett ,1830) L 

Thalassoma quinquevittatum (Lay et Bennett ,1839) L 

Weltmorella nigropinnata (Seale,1900) L 

Genus sp.287 L 

SCARIDAE 

Hipposcarus longiceps Valenciennes, 1839 L 

Scarus frenatus Lacepede,1802 L 

Scarus ghobban (Forsskal ,1775) L 

Scarus gibbus Riippell,1828 L 

Scarus globiceps Valenciennes, 1839 L 

Scarus oviceps Valenciennes, 1840 L 

Scarus psittacus Forsskal, 1775 L 

Scarus rubroviolaceus (Bleeker ,1849) L 

Scarus sordidus Forsskal, 1775 L 

Scarus sp.(venosus) L 

Scarus sp.106 (juv.) L 

Scarus sp.107 (juv.) L 

Scarus sp.329 (juv.) L 

Scarus sp.422 (juv.) L 

BLENNI IDAE 

Enchelyurus ater (Gunther ,1877) L 

Istiblennius periophthalmus (Valenciennes ,183 6) L 

Petrocirtes xestus Jordan et Seale,1906 L 

CALLIONYMIDAE 

Callionymus sp.288 L 

GOBI IDAE 

Amblygobius nocturnus Smith, 1956 L 

Amblygobius phalaena (Cuvier ,183 7) L 

Asterropterix semipunctatus Riippell ,1828 L 

Callogobius sc later i (Steindachner ,1880) L 

Eviota afalei Jordan et Seale,1905 L 

Eviota infulata Smith, 1956 L 

Fusigobius neophytus (Gunther ,1877) L 

Gnatholepis cauerensis (Ble eker ,1853) L 

Nemateleotris magnifica Fowler, 1938 

Priolepis cincta (Regan, 1908) L 

Ptereleotris evides (Jordan et Hubbs,1934) L 

Ptereleotris microlepis (Bleeker ,185 6) L 

Vanderhorstia sp.(juv.) L 

Genus sp.290 L 

Genus sp.324 L 

ACANTHURIDAE 

Acanthurus glaucopareius Cuvier, 1829 

Acanthurus nigricauda Duncker et Mohr,1929 L 



34 



Acanthurus nigroris Valenciennes, 1838 L 

Acanthurus olivaceus Bloch et Schneider ,1801 L 

Acanthurus pyroferus Kittlitz , 1834 

Acanthurus triostegus (Linne, 1758) L 

Acanthurus xanthopterus Valenciennes, 1835 L 

Acanthurus sp.(juv-) LO 

Ctenochaetus striatus (Quoy et Gaimard,1824) L 

Ctenochaetus strigosus (Bennett ,1828) L 

Naso brevirostris (Valenciennes , 1835) L 

Naso hexacanthus (Bleeker ,1855) 

Naso lituratus (Bloch et Schneider ,1801) L 

Naso unicornis (Forsskal ,1775) L 

Zanclus cornutus (Linne, 1758) L 

Zebrasoma rostratum (Giinther ,1873) 

Zebrasoma scopas (Cuvier ,183 5) L 

Zebrasoma veliferum (Bloch, 1795) L 

SIGANIDAE 

Siganus argenteus (Quoy et Gaimard ,1824) 

SCOMBRI DAE 

Katsuwonus pelamis (Linne, 1558) 
Order Pleuronectiformes 

BOTHIDAE 

Bothus mancus (Broussonnet ,1 782) L 
Order Tetraodont i formes 

BALISTIDAE 

Amanses scopas (Cuvier ,1829) 

Balistapus undulatus (Mungo Park, 179 7) 

Balistapus viridescens (Bloch et Schneider ,1801) L 

Melichthys niger (Bloch, 1786) 

Melichthys vidua (Solander ,1845) 

Pseudobalistes f lavimarginatus (Ruppell , 1828) 

Rhinecanthus aculeatus (Linne, 1758) L 

Sufflamen bursa (Bloch et Schneider ,1801) 

OSTRACIIDAE 

Ostracion cubicus Linne, 1758 L 

TETRAODONT I DAE 

Arothron hispidus (Linne, 1758) L 

Arothron meleagris (Lace'pede,1798) L 

Canthigaster bennetti ( Bleeker ,1854) L 

Canthigaster j anthinoptera (Bleeker ,185 5) L 

Canthigaster solandri (Richardson, 1844) L 



m 



in 


in 


L. 




+j 


4J 


c 




4-1 


E 


B 


e 


V) 


•H 


09 


09 


■H 


4-> 


in 


■ 


■ 


4-> 


e 


o 


09 


|H 


U 


09 


a. 


in 


■o 


3 


■ 


09 


IS 


09 


l_ 


•H 


•a 


xs 


in 


4-> 


■o 








in 


09 


09 


09 


m 


c 


in 


■M 


4J 


|H 


a 




IS 


IS 


L_ 


o 


m 
o 


Q. 


B 
O 


4- 


>4- 


+J 


V) 


A 


09 


09 


■H 


a 


1- 




l n 



01 

60 
•H 




Q o°0 


>1 m 
0.5 - 1 ■ 
0.25 - 0.5 m 
0.05 - 0.25 m 

< 0.05 mn 


<£><£> 


Corals 








°°n°o° 


V*v 


Coral I inaceae 








• • • 
. • • 


? ? 


Haliaeda 










* A* 


Molluscs 








t^ 


j j; 


Foraainifera 



l km 



Figure 3 : Grain-size and components of lagoonal and outer reefal 
sediments 




SWELL 



Figure 4: Surface circulation of lagoon waters (usual climatic conditions) 




Figure 5 : Relative variations of the lagoon water level 
between 1979 and 1981. 



1000 . 



100 



flM% 



• Hoa (south) 

a Middle 

o Pass (west) 




T>-. 



y= -Q.1 1x +2.81 
- Y = -0.16x +3.11 
y = -Q24x + 3.15 



7 



m 



Figure 6 : Light penetration in 3 different points of the lagoon. 




<U 




J= 




4-1 




!>■> 




-D 




-o 




CD 




60 




cd 




E 




cd 


• 


"O 


/^N 




cd 


' — * 


0) 


e 


S-i 




cd 


o 




CN 


cd 


' — / 


■H 




4-1 


ID 


« 


a 


> 





cd 


i—i 


CD 


in 


H 


U 




<U 


CO 


4-> 


> 


3 


ai 


o 


« 


j= 


<D 


+j 


c 


h 


(0 





o 


c 


•H 




^ 


a) 


>-l 


J= 


p 



Ei .e 



at 




j= 




4-1 




c 




o 




„ 




>-l 




cd 




M-l 




•H 




TD 




CD 




a. 




CO 







■ 


4-1 


^■n 


O 





M 


M 


4-J 


•H 


c 


a 


<u 


CO 


o 


Ph 


o 


^— ' 


x> 







cu 


.— 1 


eo 





TD 


o 


•H 




M 






CD 


t-( 


c 


a) 


•l-l 


60 


J3 


i-l 


o 


Cfl 


M 




D 


C 




u 


4J 


CD 


CI) 


XI 


£ 


4-J 


i— i 


3 


OJ 


O 


sa 


CO 



ATOLL RESEARCH BULLETIN 
No- 287 



CHECKLIST OF THE VASCULAR PLANTS OF THE 
NORTHERN LINE ISLANDS 



By 
Lyndon Wester 



Issued By 

THE SMITHSONIAN INSTITUTION 

Washington, D- C«, U.S.A. 

May 1985 



CHECKLIST OF THE VASCULAR PLANTS OF THE 
NORTHERN LINE ISLANDS 



By 



Lyndon Wester * 



INTRODUCTION 

The Northern Line Islands consists of four atolls aligned on 
an an axis which runs from south east to north west. The three southern 
islands Christmas (Kiritimati) , Fanning (Tabuaeran) and Washington 
(Teraina) have permanent populations and are part of the Republic of 
Kiribati (Table 1). The fourth island, Palmyra, on the north end of the 
chain, is an unoccupied U.S. possession. 

Table 1 
Northern Line Islands 





Land area 


Rainfall 


Pop 


iulation 


Political 




(sq. kms. ) 


(millimeters) 






jurisdict 


Palmyra 


0.6 


4161 







U.S. 


Washington 


14.2 


2902 




417 


Kiribati 


Fanning 


34.6 


2086 




434 


Kiribati 


Christmas 


363.4 


766 




1288 


Kiribati 



Sources: Carter, 1984; Ministry of Education, Training and Culture, 
1979; Taylor, 1973. 



The islands are remarkably dissimilar considering their 
proximity. This is in part due to the fact that they lie across an abrupt 
rainfall gradient. Christmas in the south, in the equatorial dry belt, 
receives only 766 mm of rain per year, whereas the islands further north 
are influenced by the intertropical convergence to a progressively 



* Department of Geography, University of Hawaii 



greater extent. Palmyra, four degrees of latitude north, receives 4161 
millimeters per year and supports a luxuriant forest. 

The islands are also quite different in form. Christmas is a 
very large atoll; most of the land is one continuous surface which 
almost completely encircles an embayment or lagoon and there is a large 
protruding peninsula off to the southeast. The island contains 
extensive plains of limestone hardpan, numerous shallow pools, beach 
berms and sand dunes. The ocean coast is characteristically a sandy 
beach. Fanning on the other hand corresponds more closely to the 
popular image of an atoll as it is made up of three long, narrow 
islands which surround a shallow lagoon. Its ocean shore is covered 
almost completely with plate-like coral shingle and little sand. 
Washington is perhaps the most peculiar island of the group. It has the 
smallest coral platform and the island is lens-shaped. Instead of a 
lagoon open to the sea the central depression of the island contains a 
freshwater lake and two peat bogs. The shore has a narrow fringing reef 
usually covered on the landward side with a thin strip of sand. 
Although Palmyra is a slightly larger coral structure than Washington, 
it is mostly submerged reef. At the time of first survey the atoll 
consisted of about fifty tiny islets heavily vegetated down to high 
tide level; however dredging and reclamation have greatly increased its 
area. 

The islands were uninhabited at the time of European discovery 
but there is ample evidence of former Polynesian occupancy (Emory, 
1934, 1939; Finney, 1958). Whalers and traders stopped at the islands 
during the early nineteenth century but the first attempt at settlement 
was by a group from Hawaii in 1820. The colony of about forty people 
included both Europeans and Hawaiians but appears to have been a 
failure because most of the party had returned by 1822 (Maria Loomis, 
1819-24; Elisha and Maria Loomis, 1820-24). Whalers who stopped by the 
island for wood or coconuts recorded an occasional castaway over the 
next two decades. However by 1840 a white man and 30 Society Islanders 
were living on the island and able to supply one of the ships engaged 
in the U.S. Exploring Exdedition with "watermelon, taro and pumpkins" 
(Anonymous, 1838-41). Two years later a whaler reported the group was 
engaged in producing coconut oil and supplied them with arrowroot 
(Hussey, 1841-45). Edward Lucett arrived on Fanning in 1846 with a 
title and the intention of establishing a coconut oil industry. He 
noted that there was a "man of Crusoe habits" on the island who had an 
Hawaiian wife and a large family of children and grandchildren and was 
engaged in the raising of pigs. What happened to the earlier colony or 
whether this represented a relict of it is not clear (Lucett, 1851). In 
1852 John English purchased the establishment and by 1854 seems to have 
expanded to Washington Island because a whaler who stopped there 
reported he was able to trade for "sweet potatoes, coconuts and 
bananas" (English, 1857; Holley, 1853-57). Washington has no safe 
anchorage and it may have been occupied only intermittently because 
when another whaler stopped there in 1861 he noted that the natives 
could provide nothing because they had only been there for a few months 
(Greene, 1860-65). English sold his interests in the islands to William 
Greig, his assistant, and George Bicknell in 1864 who switched to the 



production of copra and were responsible for extensive planting of 
coconut on both Washington and Fanning. When George Bicknell died the 
operation of the plantation passed to the Greig family who remained 
there well into the twentieth century. 

Drought-prone Christmas Island with only small natural stands 
of coconut had little to offer whalers except turtles and fish. 
Phosphate attracted guano diggers to Christmas Island in 1858 and some 
rock was also exported from Fanning Island between 1878 and 1881 but 
the Northern Line Islands were not among the major producers of 
phosphate rock. It was not until 1882 that more or less permanent 
occupation began on Christmas. Messrs Macfarlane and Henderson of 
Auckland took possession of the island in the name of their company and 
over the next few years their employees were engaged in the gathering 
of pearl shell from the lagoon and the planting of coconuts (Bailey, 
1977). 

When the British Commonwealth communications cable was 
constructed across the Pacific from Canada to Australia, a relay 
station was built on Fanning Island which operated from 1902 to 1963. 
This imposing facility, which had a permanent staff and could boast of 
a swimming pool, tennis courts and extensive gardens, enhanced the 
position of Fanning as the focus of human activity in the Line Islands. 
In 1902 Lever Brothers Ltd. acquired a lease of Christmas Island and 
financed a major coconut planting program. The Greigs and the heir of 
James Bicknell were forced to sell their interests in Washington and 
Fanning Island in 1907 although members of the Greig family remained to 
manage the plantation. The purchaser was Emmanuel Rougier who conveyed 
them to a company called Fanning Island Limited just a few years later. 
Meanwhile by 1914 he had taken over the Lever lease of Christmas Island 
and formed the Central Pacific Coconut Plantations Limited . He and 
later his nephew Paul Rougier ran the islands as a plantation. The 
Gilbert and Ellice Islands Company took over the running of the 
Christmas Island coconut plantation in 1941 after Paul Rougier became 
embroiled in criminal and political affairs and returned to France 
(Bailey, 1977). Washington and Fanning, the wetter and more productive 
islands, were acquired by Burns Philp & Co. who continued to operate 
them as coconut plantations until they sold them to the Kiribati 
government in 1983 (Republic of Kiribati, 1983). 

During the Second World War New Zealand and American troops 
were garrisoned on Christmas Island and in 1956-58 Britain used the 
island for nuclear testing. The United States used it for similar 
purposes in 1962. All devices were detonated in the atmosphere and its 
most conspicuous martial legacy is 100 kilometers of sealed road and 
impressive quantities of abandoned equipment and rusting structures. 

Palmyra Island escaped permanent settlement or significant 
modification until the United States established a military base there 
in 1940 which was eventually expanded to accommodate 6,000 personnel. 
The transformation of the island by dredging the lagoon, constructing 
causeways and building airstrips has been described in detail by Dawson 
(1959). The island was abandoned as a base in 1958 and plans to develop 
it as a plantation or a resort have come to nothing. It remains an 



uninhabited U.S. possession recovering from profound disturbance. 

In 1979 the Gilbert, Phoenix and Line Islands, formerly 
administered by the British, became Kiribati, an independent republic. 
It is the hope of the government to use the Line Islands, particularly 
Christmas, to settle people from heavily overpopulated South Tarawa, in 
the Gilbert group. However the economic prospects for development of 
the Line Islands are small, and the problem of transport and 
communication with the administrative and population center far to the 
west, is great. Improvements to the airport and the construction of an 
hotel were sponsored by the Japanese government who built and maintains 
a down-range missile tracking station on Christmas Island. A small 
tourist industry exists on Christmas based on game fishing and some 
fish are exported to Honolulu. Copra production has diminished almost 
to zero. However the government is presently engaged in a study of the 
agricultural potential of the newly acquired Washington and Fanning 
Islands. 

SUMMARY OF THE LAND FLORA 

The indigenous land flora of vascular plants consists largely 
of widespread strand and coral island species. Endemism is low, as is 
to be expected on an atoll. The only endemic species which have been 
described from the islands belong to genera which, for one reason or 
another, pose many problems for the systematist. Hence the status of 
these taxa ( Asplenium paclficum , Pandanus f anningensls , P_. hermslanus , 
four varieties of P_. f ischerianus , Portulaca f osbergil and P_. johnll ) 
is in doubt. Only nine indigenous species, out of a total of 42, occur 
on all of the islands in the group and the significant differences in 
the floras, and the character of the vegetation communities, can be 
related to rainfall (Table 2). The smaller but wetter islands (Palmyra 
and Washington) are mostly covered by closed forest and have more 
indigenous species than the much larger Christmas Island. The 
vegetation of the latter consists of either sickly coconut plantation 
or low scrub. Fanning receives sufficient rainfall to support closed 
canopy forests of Cocos , Pisonla and Pandanus but extensive tracts of 
land are inundated during high tides and these mudflats support mainly 
Lepturus grass. 

Table 2 
Origin and status of species 

Indigenous Cultivated Adventive TOTAL 
or persisting 

Palmyra 21 14 23 58 

46 20 91 

70 30 123 

25 25 69 



Washington 


25 


Fanning 


23 


Christmas 


19 



Similarity indices calculated on the basis of the entire 
indigenous flora (Table 3) show a high level of similarity between 
Palmyra and Washington and, to a slightly lesser extent, between those 
two islands and Fanning. Christmas, on the other hand, is quite 
dissimilar from Palmyra and Washington but bears considerable similarity 
to Fanning. 

Table 3 
Similarity indices: indigenous species 

Washington Fanning Christmas 

Palmyra 78.2 63.6 40.0 

Washington 62.5 36.4 

Fanning 61.9 

Calculations based on Sorensen index of similarity 

SI = number of species common to both islands x 100 

1/2 (total species on island A + total species on island B) 



The concentration of introduced species shows a distinctly 
different pattern. Fanning has many more cultivated and adventive 
species, which can probably be explained by its suitability for 
horticulture and its long history as the headquarters for the main 
plantation on the islands. Kyte (1861) remarked on the variety of crops 
which were grown on the plantation and the owners went so far as to 
bring soil from Honolulu for their gardens. Many other plants were 
imported for the extensive gardens of the Cable Station and even today a 
number of ornamental species persist despite the lack of care. 
Washington might have been equally suitable to support crops plants or 
ornamentals but it lacks a safe anchorage and so fewer introductions 
have occurred. Most of the exotic species recorded from Palmyra were 
introduced when it served as a military base. The severely disturbed 
areas are still suitable habitats for adventive species but the 
introduced plants will probably be replaced if no further human 
interference occurs. It would appear that some adventive species 
recorded in the nineteenth century have since disappeared. The guano 
digger John Arundel for example collected Achyranthes aspera and 
Asclepias curassavica on Fanning but they have not been recorded since 
and we can assume they are locally extinct. 

There is a much lower level of similarity between the 
assemblages of adventive species on the different islands (Table 4). 
Furthermore the patterns of similarity are somewaht different. The 
highest similarity is between Fanning and Washington which is probably 
because they are both wet, have experienced similar human use and indeed 
been operated as a single plantation for most of their recent history. 



Washington 


Fanning 


Christmas 


41.9 


45.3 


45.8 




56.0 


53.3 
43.6 



Table 4 
Similarity indices: adventive species 



Palmyra 

Washington 

Fanning 

The pattern of similarity between the cultivated species on 
islands shows a similar pattern to that of the adventives (Table 5). 
Fanning and Washington again show a high level of similarity furthermore 
Fanning, Washington and Christmas together as a group seem to have much 
in common but are quite dissimilar from Palmyra. This may be because 
Palmyra has had a very different history of human occupation and 
disturbance. 

Table 5 
Similarity indices: cultivated or persisting 



Washington 


Fanning 




Christmas 


Palmyra 30.0 


21.4 




35.9 


Washington 


65.5 




53.5 


Fanning 






46.3 


PLANT COLLECTORS OF THE 


NORTHERN LINE 


ISLANDS 





Of the four islands of the Northern Line Islands, Christmas 
and Fanning have received most attention from plant collectors because 
they are more accessible. A few specimens remain from collections 
gathered in the nineteenth century but the first systematic inventories 
were made by participants in expeditions sent by the Bishop Museum in 
the 1920's and 30's. Since that time there have been other efforts 
which have added one or two new indigenous species to the known flora 
and made it possible to keep track of introductions. A summary of 
information about each of the collectors who have worked in the 
Northern Line Islands, and the disposition of their specimens, has been 
compiled for reference. 

Arundel, John T. was a trader and guano digger who became one of the 
leading figures in the Pacific phosphate industry (Langdon, 1974). He 
was at first a field manager for the British firm of Houlder Bros, and 
Co. which operated in the equatorial Pacific islands. He later went 
into the business himself and, between 1883 and 1891, operated from 
Apia using mostly Niue and Cook Island laborers. At various times he 
held leases for many of the dry guano islands. He traveled extensively 
and is known to have visited the Line Islands in 1873. Between 1879 and 
1881 he directed the guano mining on Fanning (Arundel, 1870-1919). On 



one these trips he apparently collected 21 specimens on Fanning and 
other islands, which were sent to Joseph Hooker at Kew (Arundel, 1890). 
From these a list was compiled (Anonymous, 1874-86) which was reported 
in part by Hemsley (1855) in the results of the Challenger Expedition. 
His specimens are preserved at Kew. 

Ball, Stanley C . was a zoologist and Curator of Collections at the 
Bishop Museum who participated in the Fanning Island Expedition in the 
company of C.E. Edmondson. They made comprehensive biological 
collections during a ten day stay on Fanning in July and August 1922 
(Edmondson, 1923; Gregory, 1923). Ball made the collections of plants 
which, along with his field notebooks, are in the Bishop Museum. In 1924 
Ball was on Christmas and again on Fanning Island but this time in the 
company of G.P. Wilder as members of the scientific party on the "Cruise 
of the Kaimiloa" sponsored, in part, by the Bishop Museum. Ball made no 
further collections at this time. 

Bennett, Frederick Debell was the surgeon on board a whaling ship which 
circumnavigated the globe between 1833 and 1836. During the voyage he 
stopped on Christmas Island (6-10 May, 1835) and made extensive plant 
collections. A list of the plants collected was published along with his 
account of the voyage (Bennett, 1970). His specimens from this voyage 
were sent to Berlin Herbarium (Lanjouw and Staflen, 1954) and presumably 
destroyed during the Second World War. It is possible that some 
duplicates may exist at the British Museum or at Kew. 

Bergman, H.F . and Erling Christophersen were botanists on the 
Whippoorwill Expedition sent to the Line Islands by the Bishop Museum. 
Bergman was responsible for systematic collecting and made extensive 
collections while on Christmas (31 July and 7 August) and on Washington 
(13-18 August 1924) (Gregory 1925). He also visited Fanning with other 
members of the expedition but on this island Christophersen seems to 
have made all of the collections. His specimens are preserved in the 
Bishop Museum and the U.S. National Herbarium and were used in the 
preparation of a detailed report on the vegetation of the islands by 
Christophersen (1927). 

Browne, Ashley was employed by the University of Hawaii Agricultural 
Extension, and selected as a member of the official party of a ship 
dispatched to supply a group of young men from Honolulu who were living 
on the Southern Line Islands (Bryan, 1974). The ship stopped at Palmyra 
on 17 October 1939 during which time Browne collected a few specimens 
which are now at Berkeley. 

Bryan, Edwin H. was Curator of Collections at the Bishop Museum when he 
made at least two stops on Palmyra Island during the 1930's. He 
travelled with ships which transported and supplied young men from 
Honolulu who were sent to occupy the Southern Line Islands in an effort 
to strengthen the United States ' s claim to that territory. In the course 
of these voyages the ships visited the Northern Line Islands. During 
stops on Palmyra (23 March 1935 and 11-12 August 1938) Bryan took the 
opportunity to make collections of plants (Bryan, 1974). His specimens 



8 

are in the Bishop Museum and the U.S. National. However the labels 
show confusion, in some cases, about the site of collection. 

Christophersen, Erling and H.F. Bergman were the botanists on the 
Whippoorwill Expedition sent by the Bishop Museum to survey the Line 
Islands. It was the responsibility of Christophersen to study the 
ecological aspects of the islands. However he also made all the 
collections on Fanning during their stay (29-30 July) and both men 
collected while they were on Christmas (31 July to 7 August) (Gregory, 
1925). However all the collecting on Washington appears to have been 
done by Bergman. Christophersen wrote a detailed report of the 
vegetation of the Line Islands based on these observations 
(Christophersen (1927). 

Cooke, Charles Montague Jr . was a malacologist at the Bishop Museum who 
accompanied Henry Cooper and Joseph Rock on an expedition to Palmyra 
Island in 1913 (Rock 1916). He was a leader of "Trip B" of the 
Whippoorwill Expedition which visited the Line Islands again in 1924 
(Gregory, 1924); however all of their important work was done on Baker 
and Howland Islands. Cooke was the leader of the Mangarevan Expedition, 
sponsored by the Bishop Museum, which stopped at Fanning Island (20-29 
April 1934) on the way south. In the course of the return journey they 
called at Christmas Island (21-22 October) and again at Fanning (23 
October). The botanists of the party were Harold St. John and F. Raymond 
Fosberg who did most of the collecting independently of Cooke (Kondo and 
Clench, 1952). 

Cooper, Henry E. was a judge in Honolulu and President of the Board of 
Regents of the College of Hawaii. In 1913, soon after purchasing the 
island of Palmyra, he took a group of scientists on an expedition of 
exploration. Joseph Rock wrote the report of the trip (Rock, 1916). 
Cooper was listed along with CM. Cooke as a collector on that 
expedition. However Cooper collected plants independently on another 
visit to Palmyra in 1914. All specimens were given to the Bishop Museum. 

Dawson, E. Yale , a marine biologist, was on Palmyra (15-21 October 1958) 
for the purpose of studying ciguatera fish poisoning. He documented the 
considerable changes caused by the construction of a military base on 
the island during the Second World War (Dawson, 1959). His extensive 
collections of both native ruderal and cultivated species are preserved 
in the Bishop Museum and the U.S. National Herbarium. 

Fosberg, F. Raymond first visited the islands as a member of the 
Mangarevan Expedition which stopped at Fanning Island (20-29 April 1934) 
during the journey south and at Christmas Island (21-22 October) and 
again at Fanning (23 October) on the voyage home. At this time Fosberg 
was acting as an assistant to Harold St. John. Fosberg again collected 
on Christmas Island (16 August 1936) in the company of Alfred Metraux 
and his wife E.M. Metraux. His specimens are in the Bishop Museum and 
the U.S. National Herbarium. 



Gallagher, M.D . was a major in the British armed forces stationed on 
Christmas Island from June 1958 to mid June 1959 during a series of 
atomic tests. He was the founder and guiding spirit of the Natural 
History Society of Christmas Island established for the purpose of 
fostering interest in wildlife. A series of bulletins were issued which 
contained useful information about the plants and animals of the island 
(Anonymous, 1962). Major Gallagher made collections of plants which he 
sent to the Bishop Museum and published an article based on his 
observations of the birds (Gallagher, 1960). 

Hamilton, Dean C . made collections and observations of plants on the 
northern portion of Christmas Island in the vicinity of Main Camp while 
conducting an entomological survey of the island for the Plant Quarantine 
Division, Agricultural Research Service of the United States Department 
of Agriculture (11-14 April 1962). In collaboration with Alvin K. Chock, 
then of the Botany Department, Bishop Museum, a list of plants of the 
island was published in the Atoll Research Bulletin (Chock and Hamilton, 
1962). The specimens are preserved in the Bishop Muesum. 

Herms, William B . was an entomologist from the College of Agriculture of 
the University of California, Berkeley who, with Harold Kirby Jr., a 
graduate assistant, spent four months in the Line Islands investigating 
the pests of coconut. He spent most of his stay from 3 May to 27 July 
1924 on Fanning Island. However he and his assistant made a short foray 
to Washington Island (13-16 May) during which Herms was largely 
incapacitated (Herms, 1925; 1926). They made collections of plants on 
both islands. Christophersen (1927) informs us that E.D. Merrill prepared 
a manuscript of a flora of the islands based on these collections and 
that it was preserved in the Bishop Museum library. A search was made for 
this manuscript but it could not be located. However Christophersen 
further stated that he had incorporated its information into his 
published work. 

Hill, F.L . made collections on Christmas Island on 25 October 1957 and 
they are presently in the Bishop Museum. No other information about the 
collector has been found. 

Hill, Margaret was a school teacher employed by the Civil Aeronautics 
Authority during the time they maintained a base on Palmyra Island. In 
October 1949 she made a collection of 25 plants from the vicinity of the 
inhabited area of Menge islet. The plants, mostly ruderals, were 
identified by Marie C. Neal and E.H. Bryan and are preserved in the 
Bishop Museum (Dawson, 1959). 

Jenkin. R. N. and M.A. Foale conducted a study of the potential of 
Christmas Island for growing coconuts for the Directorate of Overseas 
Studies of the British Government during 1965 and 1966 (Jenkin and Foale, 
1968). They spent August and September 1965 on Christmas Island doing 
the field portion of the study and during that time Jenkin collected 
plant specimens. At least some of the specimens are at Kew. 

Judd, Albert F . was a trustee of the Bishop Museum who went as a member 
of the official party on the ship supplying groups of young men sent to 



10 

occupy the Southern Line Islands. He and D. Mitchell made collections 
while on Palmyra Island (13 June 1935) which were placed in the Bishop 
Museum. 

Kirby, Harold Jr . was a graduate student in zoology from the University 
of California who accompanied William Herms to Fanning, and presumably 
Washington Island, to study insect pests attacking the coconut. Extensive 
collections on both islands were made. They arrived at Fanning on 3 May 
and Kirby remained until 3 October although Herms left near the end of 
July. They made a short foray to Washington (13-16 May) (Herms, 1925; 
1926). Otherwise most of Kirby' s time was spent on Fanning although he 
joined the scientific party of the Whippoorwill Expedition sent by the 
Bishop Museum when they stopped on Fanning (Gregory, 1925). 

Lee, Mary Ann Bacon , a geographer from the University of Iowa, spent 
several weeks on Fanning in July 1983 to conduct a study of the effect of 
land crabs on the germination and spread of seeds. She collected plants 
mainly in the vicinity of the Cable Station and they are preserved in the 
Bishop Museum. 

Long, C.R . participated in the Pacific Ocean Biological Survey whose 
goals included an inventory of the terrestrial flora of islands of the 
Northwest Hawaiian Chain and the atolls of the Central Pacific. Long made 
two voyages to the Line Islands, during which he collected extensively. 
In the course of the first trip in 1964 he stopped on Palmyra (6-7 June), 
Washington (9-10 June) and Christmas (14-16 June) on the way south and at 
the same islands on the return trip Christmas (21-23 November), 
Washington (25-26 November), and Palmyra (27-28 November). In the 
following year on the return leg of a voyage to the southern islands he 
stopped again on Christmas Island (25-30 June) and for the first time on 
Fanning (2 July). The main set of his specimens, his collection records 
and notebook are housed in the herbarium of the Bishop Museum. There are 
in additional specimens in the U.S. National Herbarium. 

Metraux, Alfred , an anthropologist and ethnologist, along with his wife 
E.M. Metraux, was on Christmas Island (16 August 1936) with Fosberg and 
made extensive collections which are now in the Bishop Museum and U.S. 
National Herbarium. 

Mitchell, Donald P ., of Kamehameha Schools in Honolulu, travelled with 
the official party on the ship taking former students from the school who 
were sent to occupy the Southern Line Islands. He was on Palmyra Island 
(13 June 1935) and, in the company of A.F. Judd, made collections of 
plants which are preserved in the Bishop Museum (Bryan 1974). 

Moeller, Henry S . collected on Palmyra Island (28 December 1959 to 3 
January 1960) and his specimens are preserved in the Bishop Museum. 

Perry, Roger collected on Christmas (August 1979) and on Washington (June 
1979) islands. His specimens are in Kew. 

Rock, Joseph Francis Charles was a botanist at the College of Hawaii 
(which was later to become the University of Hawaii) who made important 



11 

contributions to the understanding of the flora of Hawaii and China. He 
and a zoologist from the Bishop Museum were invited by the owner of 
Palmyra, Henry E. Cooper, to accompany him on an expedition to that 
island in 1913. They made extensive collections between July 12 and 28th 
and Rock wrote a detailed description of the island illustrated by 
excellent photographs (Rock, 1916). The publication, produced in 
cooperation with several specialists, includes lists of fungi, lichens, 
mosses, ferns and higher plants as well as descriptions of new species 
and forms. Specimens are preserved in the Bishop Museum and the U.S. 
National Herbarium. Rock also wrote a popular account of the trip which 
was published in the Atlantic Monthly (Rock, 1929). 

Russell, Dennis J . and Roy T. Tsuda collected on Fanning Island in July 
1972 while they were graduate students in botany at the University of 
Hawaii. Their specimens were presented to the Bishop Museum and were 
consulted by Harold St. John when he prepared the flora of Fanning (St. 
John, 1974). 

St. John, Harold was the botanist on the Mangarevan Expedition led by C. 
Montague Cooke and sponsored by the Bishop Museum. With the assistance of 
F. Raymond Fosberg he collected on Fanning Island (20-29 April 1934) 
during the southward passage and on Christmas Island (21-22 October) and 
Fanning (23 October) on the voyage back to Honolulu (Gregory, 1935). 
Collections were made by St. John and Fosberg as well as St. John and 
Cooke. The specimens are preserved in the Bishop Museum. St. John also 
made the determinations of the specimens collected by Russell and Tsuda 
on Fanning in 1972 and prepared a flora of this island (St. John, 1974). 

Tsuda, Roy T . and Dennis J. Russell collected on Fanning in July 1972 
while they were graduate students in botany at the University of Hawaii. 
Their specimens were presented to the Bishop Museum and were consulted by 
Harold St. John in the preparation of the flora of the island (St. John, 
1974). 

Streets, Thomas H . was the assistant surgeon on board the U.S.S. 
Portsmouth, commanded by Joseph S. Skerrett, which was engaged in the 
United States North Pacific Surveying Expedition between 1873 and 1875. 
This was a hydrographic survey conducted by the U.S. Navy to check 
hazards to navigation in the Pacific and Lower California. In the course 
of this they stopped on Palmyra (12-27 December 1873), Washington (31 
December 1873 to 3 January 1874), Fanning (4 January 1874) and Christmas 
Islands (14-22 January 1874) (Skerrett, 1873-4). Streets and the surgeon, 
William H. Jones, made plant collections and gathered information on 
animal life sufficient to write three articles on the birds and natural 
history of the islands (Streets, 1876, 1877a, 1877b). Most of the plant 
specimens were sent ahead to Asa Gray who made determinations. By the 
time Streets returned from the expedition the specimens had been 
distributed through the herbarium of the Department of Agriculture. No 
complete list of the specimens had been made. The list later published by 
Streets was based on duplicates he had retained of material collected on 
Palmyra, Washington and Christmas Islands (Streets, 1877a). Some of 
Streets' specimens are in the U.S. National Herbarium. On the labels the 



12 

collector was first shown as Dall (or Dale) but this has been crossed out 
and replaced by the name Dr. Streets. 

Wester, Lyndon L . made collections during two trips to the Northern Line 
Islands. In the course of a reconnaisance of the vegeation in 1982 he 
collected on Fanning (4-11 August), Washington (6 August) and Christmas 
(12-19 August). In the following year a longer stay was made on 
Washington Island (7-21 August) in the company of James 0. Juvik and Paul 
Holthus for the purposes of conducting a vegetation survey and obtaining 
peat from the bog for pollen analysis. Transport to Washington required 
stops on Fannning Island and some additional collecting was done (6-7 and 
21-22 August). All specimens are preserved in the Bishop Museum. 

Wilder, Gerrit Pamile was an horticulturalist who, along with S. Ball, 
was a member of the scientific party on the "Cruise of the Kaimiloa". 
Wilder made plant collections during stops on Fanning (27 November to 7 
December 1923) and Christmas (8-17 December). The specimens were 
deposited in the Bishop Museum (Gregory, 1925). 



CHECKLIST 

In the list of species below the p 
other of the Northern Line Islands is indie 
(Palmyra), W (Washington), F (Fanning) or C 
by the names (abbreviated) of the person, o 
specimens consulted for this work. In some 
three collectors visited an island at the s 
both individually and in pairs or trios. Th 
collectors were not differentiated and any 
that group are designated in the same manne 
collectors and groups of collectors are ind 



13 



resence of a species on one or 
ated by the symbols P 

(Christmas). This is followed 
r persons, who have collected 
instances parties of two or 
ame time and made collections 
e various combinations of 
plants collected by members of 
r. The abbreviations for 
icated below. 



Most of the specimens taken from the Line Islands are in the 
herbarium of the Bernice P. Bishop Museum, Honolulu. The location of a 
specimen is indicated in round brackets ( ) after the collector's name 
only if it is housed in an herbarium other than the Bishop Museum. 

A few important specimens seem to be lost, or at least not found 
in the obvious place for them. They are shown in square brackets [ ] . 
Species not represented by specimens in herbaria at all, but which have 
been directly observed by the author or recorded in the literature by a 
reliable source, are designated "observed". 



Adl 


Arundel 


Bal 


Ball 


Ben 


Bennett 


Ber 


Bergman 


Brn 


Brown 


Bry 


Bryan 


Cht 


Christophers en 


Cok 


Cooke 


Cop 


Cooper 


Cur 


Curlett 


Daw 


Dawson 


F&M 


Fosberg and Metraux 


Gal 


Gallagher 


Gri 


Griggs 


Ham 


Hamilton 


H&K 


Herms and Kirby 



HiF 


Hill, F.L. 


HiM 


Hill, Margaret 


Jen 


Jenkin 


J&M 


Judd and Mitchell 


Lee 


Lee 


Lng 


Long 


Moe 


Moeller 


Pry 


Perry 


RCC 


Rock, Cooke and Cooper 


R&T 


Russell and Tsuda 


StJ 


St. John 


S&F 


St. John and Fosberg 


SFC 


St. John, Fosberg and Cooke 


Sdg 


Sledge 


Str 


Streets 


Wes 


Wester 


Wil 


Wilder 



* Introduced by Polynesians or in historic time 

[ ] Specimen was not seen. 

( ) Herbarium where the specimen is housed if other than Bishop 
Museum. 



Herbaria 

K Kew 

UC University of California, Berkeley 

US U.S. National Museum of Natural History. 



14 

PSILOTACEAE 
Psilotum nudum (L.) Beauv. 

Found mostly as an epiphyte on bases of coconut trunks. 

P - Daw, Lng(K) 

W - Ber, Pry(K), Wes 

F - R&T, Wes. 

ASPLENIACEAE 
Asplenlum nidus L. 

Holttum described A. pacificum from a plant grown at Kew. The spores 
were obtained from a specimen collected on Washington Island. However 
the status of this taxon, as distinct from A. nidus , will only be 
clear when the genus is studied more closely. It is one of the most 
common epiphytes and understory species on Palmyra and Washington. 
P - RCC, Bry(K), Bry 
W - Ber, H&K(UC), Sdg(K), Pry(K), Wes. 

NEPHROLEPIDACEAE 

Nephrolepis exaltata Schott 

Locally abundant as an epiphyte on trunks in understory. Some doubt 
exists about this species. Sledge did not give a specific name to the 
specimen he collected and F. M. Jarrett felt that the Perry specimen 
was intermediate between _E. biserrata and E_. hirsutula . 
W - Str(US), Ber, Pry(K), Wes. 

N. hirsutula Forst. 

Appears on Palmyra and shows distinct differences from the species on 
Washington. . 
P - Daw, Lng. 

POLYPODIACEAE 
Phymatodes scolopendria (Burm. f.) Ching 

Very common epiphyte and forms dense understory in coconut forest. 

Recorded as Polypodium aureum by Streets and Polypodium scolopendria 

or Microsorium scolopendria several others. 

P - RCC, J&M, Brn. 

W - Str(US), Ber(K), H&K(UC), Wes. 

F - Adl(K), Bal, Cht, H&K(UC), R&T ,Wes. 

ARAUCARIACEAE 
*Araucaria sp. 

A few large trees planted as ornamentals around the Cable Station on 

Fanning. 

F - R&T, Wes. 

1. 
PANDANACEAE 
Pandanus sp. 

A new species recognised by St. John but not yet published. 
F - STC. 

1. Fosberg (Kew Bull. 31:837-840, 1977) regards all of the Pandanus taxa 
listed here as minor taxa, cultivars, or individuals of Pandanus tectorius 
Parkinson. 



15 

P. fanningensis St. John 

A species known from only two specimens collected in 1972 near the 
Cable Station on Fanning. 
F - R&T. 
P. f ischerianus Martelli 

var. rockii (Mart.) B.C. Stone 

A specimen from Palmyra collected by Rock was described by 
Martelli (in Rock 1916) as a new species, P. rockii . However 
Stone (1968) believed this taxon is better regarded as a variety 
of P^_ fischerianus . 
P - RCC, J&M, Moe. 
var. cooperi (Mart ex Rock) B.C. Stone 

Material collected by Rock from Palmyra was described by Martelli 
(in Rock, 1916) as a new variety of P. pulposus (var. cooperi 
Mart, ex Rock). However after intensive study of the Pandanus of 
the Marshall Islands Stone (1968) concluded that this taxon is 
better regarded as a form of P. fischerianus . 
P - RCC 
*var. pulposus (Warburg) B.C. Stone forma bergmanii (F.Br.) B.C.Stone 
A specimen collected by Bergman on Washington Island was 
described by Brown (1930) as a new species, P. bergmanii F. Br.. 
Stone (1968) believed this to be a cultivated variety similar to 
some found in the Gilbert Islands and possibly introduced by 
workers. He concluded that this taxon should be regarded as a 
form of _P^ fischerianus . 
W - Ber 
var. bryanii B.C.Stone 

A specimen collected by Bryan on Palmyra in 1935 was described as 
a new variety of P. fischerianus by Stone (1968). St. John (1983, 
pers. comm.) believes that this taxa should be raised to the 
species level but he has not published the new name. This wild 
species has also been collected on Washington Island. 
P - Bry. 
W - Wes. 
P. hermsianus Mart. 

A single phalange collected on Fanning Island by Herms was the basis 
upon which Martelli (1926) described the species P. hermsianus Mart. 
He believed the phalange had drifted from elsewhere and that the 
species was not native to Fanning. Stone (1968) thought there was 
insufficient material to create a new species but St. John (1972) 
concluded that a specimen he and Fosberg collected on Fanning in 1934 
belonged to this taxon and was able to provide a more complete 
description. 
F - H&K(UC), S&F. 
*P. tectorius Parkinson 

var. nova-caledonicus Mart. 

St. John (1972) believes this to be a cultivated species 
introduced to Fanning Island by Gilbertese laborers. Furthermore 
he thinks the specimen collected by Long in 1965 is the same as 
one photographed by Herms in 1924. 
F - Lng 



16 

POTAMOGETONACEAE 

Potamogeton sp. 

A sterile specimen, said to have been collected by Bergman in the 
lake of Washington Island (Christophersen, 1927). 
W - [Ber] . 

POACEAE 

*Cenchrus echinatus L. 

A common grass on atolls but perhaps not native. 

P - HiM. 

W - Wes. 

F - S&F, R&T, Wes. 

C - Gal, Ham, Wes. 
*Chloris inflata Link 

An adventive on Palmyra in 1949 (Dawson, 1959). 

P - HiM 
*Cynodon dactylon (L.) Pers. 

A common lawn species on Washington and Fanning. 

W - Ber, Wes. 

F - H&K(UC), Wes. 
*Dactyloctenium aegyptium (L.) Willd. 

An uncommon weed in waste areas around Napia village on Fanning. 

Perhaps a new arrival. 

F - Wes. 
Digitaria pacif lea Stapf 

This is the Syntherisma pelagica F. Brown (variety b) which was 

described in Brown (1931) and the plant identified by Christophersen 

(1927) as Panicum stenotaphrodes Nees. ex Stend. 

C - Ber, SFC, F&M, Gal, Lng, Wes. 
*Digitaria sp. 

In grassy areas around village on Washington. Said by one of the 

residents to be a new arrival. 

W - Wes. 
*Eleusine indica (L.) Gaertn. 

A common volunteer in waste places. 

P - Daw. 

W - Ber, Wes. 

F - S&F, R&T, Wes. 

C - Gal, Ham, Lng, Wes. 
*Eragrostis ciliaris (L.) R.Br. 

Rare in waste areas. 

C - Wes. 
*E. pilosa (L.) Beauv. 

Rare in waste areas 

C - Wes. 
*E. tenella (L.) Beauv. ex R.& S. 

Recorded as E_. amabilis (L.) Wight and Arnott by Christophersen 

(1927) and Chock & Hamilton (1962). A common weed. 

W - Ber, Wes. 

F - H&K(UC), R&T, Wes. 

C - Ber, Gal, Wes. 



17 

E. whitneyi Fosb. 

Listed as E. falcata (Gaud.) Gaud, by Christophersen (1927) (See 

Fosberg, 1939) 

C - Ber, SFC, F&M, Lng, 
Lepturus repens (Forst. f.) R.Br. 

Commom in natural and open areas, along roads and in understory where 

shading is not excessive. This was designated as "Haemoenthuia 

conf itessa " on the list of plants collected by Arundel. 

P - RCC(K), Bry, J&M, Daw, Lng. 

W - Ber, Lng, Wes. 

F - Adl(K), Bal, Cht, Wil, H&K(UC), Lng, R&T, Wes. 

C - Ber, SFC, F&M, Gal, Ham, Lng, HiF(K), Wes. 
*Panicum maximum Jacq. 

Misidentif ied as P. barbinode Trin. 

C - S&F. 
*Paspalum f imbriatum H . B . K . 

Dawson (1959) found this naturalized on Cooper islet of Palmyra. 

P - Daw. 
*P. orbiculare Forst. f. 

Dawson (1959) found this naturalized on Menge islet of Palmyra. 

P - Daw. 
*Rhynchelytrum repens (Willd.) C.E.Hubb. 

Also known as Tricholaena rosea Nees. Small colony perpetuating 

itself around Fanning Is. Cable Station. 

F - R&T, Wes. 
*Saccharum officinarum L. 

Cultivated in village on Washington. 

W - observed. 
*Sporobolus indicus (L.) R. Br. 

= £. poiretii (R. & S.) Hitch. 

On Palmyra Dawson (1959) found naturalized on Menge islet and in 

disturbed area on Cooper Island and has characteristic large, almost 

oblong seeds. Another specimen of Sporobolus is in the Bishop Museum 

with a notation on the label which reads 'could be from Palmyra 

according to Bryan". 

P - Daw, Lng. 
*Stenotaphrum secundatum (Walt. ) 0. Kuntze 

Planted as a lawn around Cable Station on Fanning but is spreading 

somewhat in to waste areas. Included in St. John (1972) list as 

Brachiaria plantaginea . 

F - R&T, Wes. 

CYPERACEAE 
*Cyperus compressus L. 

A few individuals found in waste area near airport terminal. Perhaps 
a new introduction. 
C - Wes. 
£. javanicus Houtt. 

A conspicuous but uncommon sedge on Washington found mainly near wier 

and in disturbed areas. Also listed as C. pennatus Lamarck. 

P - Daw. 

W - Ber, Sdg(K), Wes. 



*C. kyllingia Endl. 

A small colony found in grassy area of village on Washington. Perhaps 

a new introduction. 

W - Wes. 
C_. polystachyos Rottb. 

A commom sedge in standing water at fringes of bog and on elevated 

mounds in bogs of Washington. 

P - HiM, Daw, Lng. 

W - Ber, Lng, Sdg(K), Wes. 
*C. rotundus L. 

An uncommon sedge found near habitations. 

F - R&T, Wes. 

C - SFC, F&M, Sdg(K). 
Fimbristylis atollensis St. John 

A common sedge found extensively in dry open natural sites and in 

waste areas around human habitations. Often combined with _F. cymosa 

R. Br. This is the species which Christophersen listed as 

F_. spathacea Roth. 

P - HiM, Daw, Lng. 

W - Ber, Lng, Wes. 

F - Bal, Cht, Wil, H&K(UC), R&T, Wes. 

C - Ham, Lng, Wes. 
Scirpus littoralis Schrader 

The dominant species over most of the bog on Washington. Also 

determined as j>. riparius Presl by Streets and Christophersen. 

W - Ber, Sdg(K), Wes. 

ARECACEAE 
*Cocos nucifera L. 

Reported in earliest accounts of all the islands but it may have been 

an aboriginal introduction. Not represented in herbarium collections. 

P - observed 

W - observed 

F - observed 

C - observed 
*Phoenix dactylif era L. 

A few individuals grown in cultivation on Fanning. 

F - observed by Wes. 
*Livistona chinensis (Jacq. ) Mart. 

A single specimen by main building of Cable Station on Fanning. 

F - observed by Wes. 

ARACEAE 
*Colocasia esculenta (L.) Schott 

Anonymous, 1940, Keyte (1861) and Bryan (1942) reported seeing it in 

cultivation on Fanning. 

F - observed 
*Cyrtosperma chamissonis (Schott.) Merr. 

Cultivated on Fanning and Washington for food but also naturalized or 

persisting in bog on Washington. This could also be the "ape " 

reported by Judd (1859). 

W - Ber, Wes. 

F - observed 



19 

*Scindapsus aureus (Linden ex Andre) Engl. 

= Epipremnum aureum (Linden ex Andre) Bunting. 

It was introduced as an ornamental to Palmyra but has become locally- 
naturalized. 
P - Daw. 

BROMELIACEAE 
* Ananas comosus (L.) Merr. 

Bryan (1942) reported seeing it in cultivation on Fanning. 
F - observed. 

COMMELINACEAE 
*Rhoeo spathacea (Sw.) Stearn 

A cultivated ornamental on Fanning. 
F - R&T, Wes. 

LILIACEAE 
*Cordyline fruticosa (L.) Chev. 

A cultivated species also listed as C_. terminalis (L.) Knuth. 

W - Wes. 
*Gloriosa superba L. 

A cultivated ornamental which persists in waste places around Cable 

Station on Fanning and on Washington. 

W - Wes. 

F - R&T, Wes. 

AMARYLLIDACEAE 
* Agave sisalana Perrine ex. Engelm. 

A few individuals were observed on Fanning by wharf at Cartwright 

Point on north side of main pass. 

F - observed 
*Crinum amabile Donn 

Cultivated around Cable Station. Also known as C_. augustum Roxb. and 

£. procerum Herbert and Carey. More material is needed of this plant 

for study. 

F - Wes. 
*C. asiaticum L. 

A robust species found in cultivation on Washington, Fanning and 

Christmas. 

W - Wes. 

F - observed 

C - observed 
*C. bulbispermum (Burm. f.) Milne-R. & Schw. 

Cultivated around Cable Station on Fanning. 

F - Wes. 
*Hymenocallis littoralis (Jacq. ) Salisb. 

Cultivated species seen around Cable Station on Fanning. Also known 

as Pancratium littorale Jacq. 

F - Wes. 
*Zephyranthes grandif lora Lindl. 

Cultivated species on Fanning and Washington which appears to have 

escaped into waste areas on Washington. 

W - Wes. 

F - Lng, Wes. 



20 

TACCACEAE 
*Tacca leontopetaloides (L.) Ktze. 

This cultivated species which was observed by Hussey (1841-45), 
Lucett, (1851), and Bryan (1942) on Fanning. It grows wild on many 
atolls and could be an aboriginal introduction or may have been 
brought by the early settlers. It persists in abandoned gardens near 
Cable Station on Fanning. 
F - Wes. 

MUSACEAE 
*Musa paradisiaca L. 

This cultivated species was recorded on Washington at least from 1854 
(Holley, 1853-57). 
W - observed by Wes. 
F - observed by Wes. 
C - observed by Wes. 

CANNACEAE 
*Canna glauca L. 

Cultivated in garden of plantation manager on Fanning in 1983. 
F - observed by Wes. 

CASUARINACEAE 
•Casuarina equlsetif olla L. 

Cultivated trees recorded from all islands. 

P - Daw. 

W - Wes. 

F - R&T, Wes. 

C - Lng, Wes. 

MORACEAE 
*Artocarpus altilis (Parkins.) Fosb. 

In cultivation on Fanning, Washington and Christmas but groves were 

observed on Washington in remote areas which seemed to be reproducing 

naturally. 

W - observed by Wes. 

F - R&T, Lng. 

C - observed by Wes. 
*Ficus carica L. 

A cultivated tree on Washington and Fanning. 

W - Wes. 

F - Adl(K). 
*F. prolixa Forst f. 

Large, mature trees found along roads and at sites of former camps on 

Washington. 

W - Wes. 

F - Adl(K), Wes. 
*F. tinctoria Forst. f. 

In cultivation around settlements. 

F - Wes. 

C - Wes. 



21 

URTICACEAE 
*Pilea microphylla (L.) Liebm. 

A widespread naturalized species on Palmyra 
P - HiM, Daw, Lng. 
Laportea ruderalis (Forst. f.) Chew 

Rock (1916) reported it to be abundant on Palmyra at the time of his 

visit. Common on Fanning and Washington usually in open areas often 

far from habitations. Also recorded as Fleurya ruderalis (Forst. f.) 

Gaud, ex Wedd. 

P - RCC, Bry, J&M, Lng. 

W - Ber, Lng, Sdg(K), Wes. 

F - Adl, Bal, Cht, H&K(UC), S&F, Lng, R&T, Lee, Wes. 

Pipturus argenteus (Forst. f.) Wedd. 

An understory shrub where canopy is open and a colonist in cleared 

areas. 

W - Lng, Sdg(K), Wes. 

POLYGONACEAE 
*Antigonon leptopus H.& A. 

Cultivated in garden around Cable Station on Fanning. 

F - Wes. 
*Coccoloba uvifera (L. ) L. 

A cultivated tree observed by Dawson (1959) on Menge, Marine Engineer 

and Cooper islets of Palmyra where it appears to be naturalized or 

persisting. 

P - HiM, Daw. 

AMARANTHACEAE 
*Achyranthes aspera L. 

= A. indica (L.) Mill. 

Presumed to be a naturalized species collected by Arundel but has not 

been recorded since. There appears to be some confusion between this 

species and _A. indica (L. ) Mill. 

F - Adl(K). 
*Amaranthus viridis L. 

An uncommon naturalized herb. 

W - Ber, Wes. 

NYCTAGINACEAE 
Boerhavia tetrandra Forst. f. 

Very common on Christmas but present in open areas and disturbed 

sites on all islands. The taxonomy of this species or group of 

species needs attention. B. repens L. sensu lato may also be present 

and the name _B. diffusa L. has been misapplied to some specimens from 

these islands. 

P - RCC, Bry,J&M(K), Daw, Lng. 

W - Ber, Lng, Wes. 

F - Bal, Wil, H&K(UC), Lng, R&T, Wes. 

C - Ber(K), Lng, Cur(K) , Gri(K), Jen(K), Sdg(K), Wes. 



22 

*Bougainvillea sp. 

Cultivated and persisting in abandoned gardens. It is unclear to me 

which species is present. 

W - Wes. 

F - Wes. 

C - observed 
*Mirabilis jalapa L. 

Found cultivated and as an escape in waste areas near settlements. 

W - Wes. 

F - R&T, 

C - Lng. 
Pisonia grand is R.Br. 

Forms spendid forests on Fanning, Washington and Palmyra and a few 

small groves on Christmas. 

P - RCC, Cop, Bry, Daw, Lng. 

W - Ber, Wes. 

F - Wil, H&K(UC), S&F, Lng, Wes. 

C - Wes. 

AIZOACEAE 
Sesuvlum portulacastrum (L.) L. var. griseum Deg. and Fosb. 

Mat forming species found in areas subject to flooding and high 

salinity. 

F - Adl(K), H&K(UC), R&T, Wes. 

C - Str(US), SFC, Gal, Ham, Lng, Wes. 

PORTULACACEAE 

P_. johnii v. Poelln. 

A specimen collected by St. John and Cooke from a small island in the 
lagoon of Christams Island was among the specimens consulted by von 
Poellnitz when he described the species (v. Poellnitz, 1936) 
C - SFC 

P_. lutea Soland. ex Forst. f . 

Very common on Christmas and in open areas on Fanning. 
F - Cht, H&K(UC), R&T, Lee, Sdg(K), Wes. 
C - Ber, Wil, SFC, F&M, Gal, Ham, Lng, Wes. 
*P. oleracea L. 

A cosmopolitan species which, in the Line Islands as elsewhere, is 
commonly found along roadsides and waste areas. Von Poellnitz (1936) 
recognized two closely related species, P_. f osbergii v. Poelln. and 
P. johnii v. Poelln., which colonized natural habitats. Fosberg 
Tl943) speculated that P_. f osbergii was intermediate between P. 
oleracea and P_. lutea but thought that further study of living 
material was needed to establish the relationship betweem these 
species. Geesink (1969) reduced both P. f osbergii and P. johnii to 
synonyms. Rock (1916) noted a single plant of what he called P_. 
oleracea on Holei Island of Palmyra atoll but appears not to have 
collected it. Dawson (1959) found P. oleracea (which he listed as P. 
f osberii ) which is probably the same species as Rock saw. 
P - Daw. 
W - Wes. 

F - Adl(K), R&T, Wes. 
C - F&M, Gal, Ham, Lng, Wes. 



23 

ANNONACEAE 
*Annona squamosa L. 

A cultivated species on Fanning collected by Bergman and also noted 
by Bryan (1942). 
F - Ber. 

LAURACEAE 
Cassytha f iliformis L. 

Very common on Christmas and found in a few open habitats on Fanning 

and Washington. 

W - Ber, Wes 

F - H&K(UC), R&T. 

C - Ber, F&M,Gal, Ham, Wes. 

HERNANDIACEAE 
*Hernand_ia sonora L. 

= H_. nymphaeif olia (Presl) Kub. 

This provisional determination is based on sterile material. A few 

individuals found cultivated on Washington. 

W - Wes. 

CRUCIFERAE 
*Brassic_a oleracea L. 

Cultivated in gardens on Fanning from an early date (Keyte, 1861) and 

also seen on Washington and Christmas. 

W - Wes. 

F - observed by Wes. 

C - observed by Wes. 

Lepidium bidentatum Mont. 

Locally common in open habitats and in some artifical clearings. Also 

recorded as L_. owaihiense C. & S. and L. piscidium Forst. 

P - Str(US), RCC, J&M, Bry. 

W - Ber, Lng, Wes. 

F - Adl(K), Cht, H&K(UC), S&F, Lng, R&T, Wes. 

C - Cht, Lng. 

CRASSULACEAE 
*Kalanchoe pinnata (Lam.) Pers. 

A cultivated species which persists in abandoned gardens. 
F - Long, R&T, Wes. 

FABACEAE 
*Bauhinia monandra Kurz 

St. John (1972) reported that this species was collected by Long but 

the specimen can not be located. 

F - observed 
*Caesalpinia pulcherrima (L. ) Sw. 

A cultivated shrub observed by Russell and Tsuda on Fanning (St. 

John, 1972). 

F - observed 



24 

Canavalia carthartica Thouars 

Common strand species on Washington. This is the £. microcarpa (DC.) 

Piper of Christophersen (1927) and probably the £. grandif lora 

recorded by Streets (1877). 

W - [Str], Ber, H&K(UC), Lng, Sdg(K), Wes. 
* Cassia occidentalis L. 

A volunteer in open areas around settlements. 

F - Adl(K), H&K(UC), Wes. 
*Crotalaria incana L. 

Locally naturalized on Palmyra. 

P - Daw, Lng. 
*C. retusa L. 

Locally naturalized around Napari on Fanning. 

F - Wes. 
*Crotalaria sp. 

One sterile specimen found in waste area on Washington. 

W - Wes. 
*Delonix regla (Boj.) Raf. 

Ornamental in village on Washington. 

W - Wes. 

*Desmodium trif lorum (L.) DC. 

Reported to grow along paths in coconut groves by Christophersen 

(1927). 

F - H&K(UC). 
*Erythrina variegata L. var. orientalis (L. ) Merr. 

Christophersen (1927) reported E. indica was grown as an ornamental 

on Christmas and Bryan (1942) also observed an Erythrina on that 

island growing around settlements. 

C - observed 
*Leucaena leucocephala (Lam.) De Wit 

Occasionally naturalized in and around settlements. Also reported as 

L_. glauca sensu Hawn. bot. non (L.) Benth. 

P - Daw. 

W - Lng, Sdg(K), Wes. 

F - R&T, Wes. 

C - SFC, F&M, HiF. 
*Peltophorum pterocarpum (DC. ) Backer ex K. Heyne 

Large ornamental tree near Cable Station. Sterile specimen 

misidentif ied as Jacaranda acutifolia Humb. & Bonpl. (St. John, 

1972). 

F - R&T, Wes. 
*Phaseolus lathyroides L. 

One record from Paris on Christmas Island. 

C - F&M. 
*Trifolium sp. 

Sterile plant found in lawn of plantation manager's house on Fanning. 

F - Wes. 
*Vigna luteola Benth. in Mart. 

This plant, which is rare in the Pacific, was found naturalized 

around Napia village of Fanning Island and was determined by Fosberg. 

F - Wes. 



25 

ZYGOPHYLLACEAE 



Tribulus cistoides L. 



Common creeping herb on Christmas. 
C - Cht, H1F, Gal, Ham, Wes. 

RUTACEAE 
*Citrus aurantiifolia (Christm.) Swingle 

Cultivated in on Washington and Fanning. 

W - Wes. 

F - observed 

S IMAROUBACEAE 
Suriana maritima L. 



Common shrub on saline soils on Christmas. Streets (1877a) indicates 
specimens were collected from Christmas and Palmyra and the plant was 
"common on all the islands of the Fanning Group" although this is 
doubtful. One small colony was noted on Fanning in 1982. Small 
populations may have been missed by recent collectors on Washington 
and Palmyra or it may be that the species is periodically 
exterminated but is able to recolonize these two islands. 
P - [Str]. 
F - Wes. 
C - Str(US), Ber, Wil, SFC, F&M, Ham, Lng, Sdg, Wes. 

EUPHORBIACEAE 
*Acalypha wilkesiana Muell. -Arg. in A. DC. 

A cultivated shrub in settlements. 

P - Lng. 

W - Wes. 

F - R&T, Wes. 
*Breynia disticha Forst. f. 

Cultivated shrub on Fanning. 

F - Wes. 
*Codiaeum variegatum (L.) Bl. var. pictum (Lodd.) Muell. -Arg. 

Cultivated shrub in villages. 

W - Wes. 

F - R&T, Wes. 
*Euphorbia glomerifera (Millsp.) L.C.Wheeler 

Introduced weed on Palmyra (Dawson, 1959). One plant seen in waste 

area around village on Christmas. The species identified as _E. atoto 

Forst. f. which appears on the Hill list (Dawson, 1959) is also 

believed to be JS. glomerifera . 

P - HiF, Daw. 

C - Wes. 
*E. heterophylla L. var. cyathophora (Murr.) Griseb. 

Common around Cable Station on Fanning and in disturbed areas on 

other islands. Also known as E. cyathophora Murr. 

P - HiM, Daw. 

W - Wes. 

F - R&T, Wes. 

C - Wes. 



26 

*E. hirta L. 

Common weed in waste areas and along roadsides. 

P - [HiM]. 

W - Ber, H&K(UC), Wes. 

F - Bal, H&K(UC), R&T, Wes. 

C - Ber, F&M, Sdg(K), Pry(K), Wes. 
*E. prostrata Ait. 

Common weed in waste areas and heavily disturbed sites. 

W - Ber, Sdg(K), Wes. 

F - H&K(UC), Lee. 

C - Wes. 
*Manihot esculenta Crantz 

Cultivated in villages. 

W - Wes. 

F - R&T. 
*Phyllanthus amarus Schum. 

Common weed misidentif ied as P. niruri L. (Christophersen, 1927) and 

as P. debilis Klein ex Willd. (Dawson, 1959). 

P - Daw. 

W - Ber, H&K(UC), Wes. 

F - Bal, Cht, H&K(UC), R&T, Wes. 

C - Ber, F&M. 

ANACARDIACEAE 
*Mangif era indica L. 

Cultivated in settlements on Fanning and Washington. 

W - Wes. 

F - observed by Wes. 

TILIACEAE 
Triumfetta procumbens Forst. f. 

Native to Polynesia, Micronesia and Malaya (Neal 1965) and used for 

fiber, ornament, magic and medicine (Luomala, 1953). Probably native 

to the Northern Line Islands but conceivably a human introduction. 

P - Daw, Lng. 

W - Lng. 

F - Adl(K), Bal, H&K(UC), R&T, Lee, Wes. 

MALVACEAE 
*Abutilon albescens Mi q . 

Locally abundant on Christmas. Wrongly identified as A. indicum Sweet 

(Fosberg, 1943). 

C - F&M. 
*Hibiscus rosa-sinensis L. 

Cultivated and persisting around settlements on Washington, Fanning 

and Christmas. 

W - Wes. 

F - Wes. 

C - observed by Wes. 



27 

*Hibiscus tiliaceus L. 

Cultivated and escaped around settlements. 

P - HiM, Daw. 

W - Wes. 

C - Ber, F&M, Ham, Wes. 
*Malvastrum coromandelianum (L.) Garcke. 

Naturalized in waste areas. 

W - Ber, Wes. 

F - H&K(UC), Wes. 
Sida f allax Walp. 

One of the most common shrubs on Christmas. Streets (1877a) recorded 

it as S_. dielli Gray, which is probably the same, and Christophersen 

(1927) misidentif ied it as S_. cordif olia . 

F - Adl(K), Bal(K), H&K(UC), R&T. 

C - Str(US), HiF, Wes. 
*S_. rhombifolia L. 

In disturbed habitats around settlements. 

W - Ber, Wes. 

C - Ber. 

STERCULIACEAE 
*Waltheria indica L. 

Also known as W. americana L. One colony found near Cable Station, 

perhaps a new arrival. 

F - Wes. 

GUTTIFERAE 
*Calophyllum inophyllum L. 

Cultivated trees around villages. 

P - Daw. 

W - Wes. 

F - Cht, Wes. 

PASSIFLORACEAE 
*Passif lora foetida L. 

Weed around Cable Station on Fanning. 
F - R&T, Wes. 

CARICACEAE 
*Carica papaya L. 

Cultivated in settlements of Washington, Fanning and Christmas and 

mostly used for pig food. 

W - Wes. 

F - Wes. 

C - observed by Wes. 

COMBRETACEAE 
*Terminalia catappa L. 

Cultivated around settlements. 

P - Daw. 

W - Wes. 

F - Wes. 

C - observed by Wes. 



28 

MYRTACEAE 
*Psidium guajava L. 

Cultivated around settlements. 
W - Ber, Wes. 
F - Wes. 

ONAGRACEAE 
Ludwigia octovalvis (Jacq.) Raven 

Found in flooded substrate near lake on Washington. 
P - Him, Daw. 
W - Lng, Wes. 

ARALIACEAE 
*Polyscias fruticosa (L.) Harms 

Cultivated in gardens. 

W - Wes. 
*P. guilfoylei (Bull) Bailey 

Cultivated in gardens. 

F - S&F, R&T, Wes. 
*P. Scutellaria (Burm. f.) Fosberg 

Cultivated in gardens. 

W - Wes. 

F - R&T, Wes. 

OLEACEAE 
*Ligustrum sp. 

The specimen collected by Russell and Tsuda (St. John, 1972) can not 
be located in the Bishop Museum. It may have been redetermined or 
lost. 



F - [R&T] 



APOCYNACEAE 



*Nerium oleander L. 

Cultivated and persisting in abandoned gardens. 

F - R&T, Wes. 

C - observed by Wes. 
Ochrosia oppositif olia (Lam.) K. Schum. 

= Neisosperma oppositif olia (Lam.) Fosb. and Sachet. 

Found only near west end of Holei islet of Palmyra. 

P - Him, Daw. 
*Plumeria obtusa L. 

Cultivated as ornamental and used in leis. Observed on Washington, 

Fanning and Christmas. 

W - Wes. 

F - observed by Wes. 

C - observed by Wes. 
*P. rubra L. forma rubra 

Cultivated as ornamental and used in leis. 

W - Wes. 

F - observed by Wes. 

C - observed by Wes. 



29 

(P. rubra ) 

forma acutifolia (Poir.) Woodson 

Cultivated as ornamental and used in leis. Probably also on 

Christmas . 

W - Wes. 

F - Wes. 

ASCLEPIADACEAE 
*Asclepias curassavica L. 

An introduced weed which was collected by Arundel last century but 
not recorded since. 
F - Adl(K). 

CONVOLVULACEAE 
Cuscuta campestris Yuncker 

Common at South East Point of Christmas Island and occasionally 

between small ponds (Garnett, 1981). 

C - Lng, Wes. 
*Ipomoea batatas (L.) Poir. 

Cultivated in village on Washington in 1854 (Holley, 1853-57). Now 

found cultivated and growing along nearby roadsides. 

W - Wes. 
I_. pes-caprae ssp. brasiliensis (L. ) v. Ooststr. 

A pioneer on beaches and in open sites. 

P - Daw, Lng. 

W - Ber, H&K(UC), Wes. 

F - R&T, Wes. 
I_. macrantha R. & S. 

Recorded also as I_. tuba (Schlecht.) G. Don, I_. grandif lora (Choisy) 

Hall f . and I_. glaberrima Bojer. A common vine on Fanning and locally 

abundant on Washington. 

P - Bry, Daw. 

W - Ber, Wes. 

F - S&F, Lng, R&T, Wes. 
*Merremia dissecta Hallier 

In Cable Station garden on Fanning and identified by the author. 

F - Wes. 

BORAGINACEAE 
*Cord_ia sebestena L. 

Cultivated tree in London village on Christmas. 

C - Wes. 
C_. subcordata Lam. 

In scattered locations usually not far from beach. On Washington it 

seemed to be associated with fresh water seeps. 

W - Ber, Wes. 

F - Bal, R&T, Lee, Wes. 
Heliotropium anomalum (H. & A.) var. mediale Johnst. 

Very common on Christmas an in open sites on Fanning. 

F - Adl(K), Bal, Cht, Wil, H&K(UC), S&F, R&T, Sdg(K) , Lee, Wes. 

C - Ber, Wil, SFC, F&M, Gal, Ham, Jen, Cur, HiF, Gri, Wes. 



30 

Toumefortia argentea L. f. 

Typically a string of these trees is found along the top of the 

beach. Also recorded as Messerschmidia argentea (L.f.) Johnst. 

P - RCC, Bry, J&M. 

W - Ber, H&K(UC), Wes. 

F - Bal, H&K(UC), Wes, 

C - Ber, SFC, F&M, Ham, Lng, Wes. 

VERBENACEAE 
*Clerodendrum inerme (L.) Gaertn. 

A favored ornamental grown as a hedge and for its flowers which are 

used in leis. It has become established at a number of sites in the 

bog on Washington. 

W - Sdg(K), Wes. 

F - S&F, Lng, R&T, Sdg(K), Wes. 
*Lantana camara L. 

Cultivated in villages where flowers are used in leis. 

W - Wes. 

F - R&T. 
*Premna obtusifolia R. Br. 

Cultivated in village on Washington. 

W - Wes. 
*Stachytarpheta urtlcaefolia (Salisb. ) Sims 

Apparently common on Palmyra and found once on Fanning. 

P - HiM, Daw, Lng. 

F - R&T. 
*Vitex trifolia L. 

Cultivated around Cable Station on Fanning and persists on Palmyra 

and is the same as V^. negundo var. bicolor (Willd.) H. Lam (Dawson, 

1959). 

P - HiM, Daw. 

F - Wes. 

LABIATAE 
*0cimum basilicum L. 

Cultivated in gardens 

F - S&F. 

SOLANACEAE 
* Capsicum annuum L. 

Cultivated in settlements. Persists for a while after garden has been 

abandoned . 

W - observed by Wes. 

F - Wes. 

C - observed by Wes. 
*Lycopersicon esculentum Mi 1 1 . 

= Solanum lycopersicum L. 

Cultivated on Fanning and Christmas but volunteers were noticed in 

waste areas on Christmas. 

F - observed by Wes. 

C - Lng, Wes. 



31 

*Nicotiana tabacum L. 

Cultivated and able to resist attacks by land crabs. 

F - Adl(K). 

C - observed by Wes. 
*Physalis minima L. 

A volunteer in disturbed areas; determined by D. Symon. 

W - Wes. 

C - Wes. 

SCROPHULARIACEAE 
*Russelia equisetif ormis Schlecht. & Cham. 

A cultivated species which is able to persist after abandonment of 

garden. 

W - Wes. 

F - Lng, R&T, Wes. 

BIGNONIACEAE 
*Spathodea campanulata Beauv. 

Cultivated tree on Washington and Fanning. 

W - Wes. 

F - observed by Wes. 
*Tecoma stans (L.) Juss. ex B.& H. 

Also known as Stenolobium stans (L. ) D. Don and prized for its 

flowers. 

W - Wes. 

F - R&T, Wes. 

ACANTHACEAE 
*Blechum brownei Jus s . 

Dawson (1959) recorded the species Blechnum brownei . This is assumed 

to be a typographical error since no fern by this name can be found 

and the species was listed with other Acanthaceae. The specimen could 

not be located in the Bishop Museum where Dawson's specimens are 

preserved. 

P - [ Daw] . 
*Graptophyllum pictum (L.) Nees ex Griff. 

Said to be persisting on Palmyra (Dawson, 1959). 

P - [Daw] 
*Pseuderanthemum carruthersii (Seem.) Guillaum. 

A common cultivated ornamental around settlements. 

P - Daw. 

W - Wes. 

F - Lee, Wes. 

C - Wes. 

RUBIACEAE 
*Spermacoce assurgens R . & P . 

Also listed as S_. suff rutescens Jacq. and misidentif ied as Borreria 

laevis (Lam.) Griseb. A common weed in disturbed areas and along 

roadsides. 

P - Daw. 

W - Sdg(K), Wes. 

F - S&F, Lng, R&T, Wes. 



32 

♦Gardenia taitensis DC . 

Cultivated on Washington and Fanning but not common. 

W - Wes. 

F - observed by Wes. 
*Guettarda speciosa L. 

Cultivated around villages for fragrant flowers which are used in 

leis. The plant appears to volunteer around margins of settlements. 

It was once collected on Palmyra (Dawson, 1959). Could be native to 

the Line Islands. 

P - [HiM]. 

W - Wes. 

F - Wil, S&F, R&T, Lee, Wes. 

C - Lng. 
Hedyotis romanzoff iensis (C.& S.) Fosb. 

A few populations found on Christmas in vicinity of ponds and 

lagoons . 

C - Ber, SFC, F&M, Lng, Wes. 
*Morinda citrifolia L. 

Found mostly in disturbed areas around villages and a few natural 

sites. Probably introduced but could be native to these islands. 

W - Ber, Lng, Wes. 

F - Bal, R&T, Wes. 

C - observed by Wes. 

CUCURBITACEAE 
*Curcubita pepo L. 

Reported to be cultivated on Fanning in 1840 (Anonymous, 1838-41). 

Now found in villages on Washington, Fanning and Christmas. 

W - observed by Wes. 

F - Wes. 

C - observed by Wes. 
*Citrullus lanatus Schrad. in Ecklon and Zeyher. 

Reported on Fanning in 1840 (Anonymous, 1838-41). 

F - observed by Wes. 

GOODENIACEAE 
Scaevola sericea Vahl 



Recorded also by synonyms S^ taccada (Gaertn.) Roxb. and S. koenigli 

Vahl. Streets misidentif ied it as £. plumieri (L. ) Vahl. This common 

strand shrub forms dense thickets on Fanning and Christmas Island. It 

occurs in a few patches on Washington and is a colonist of disturbed 

habitats on Palmyra (Maragos, 1979). 

P - J&M, Bry, HiM. 

W - Ber, Wes. 

F - Bal, H&K(UC), S&F, Wes. 

C - Str(US), Ber, SFC, F&M, Gal, Ham, Wes. 

ASTERACEAE 
*Bidens pilosa L. 

A weed in disturbed areas around settlements. 

W - Lng, Wes. 

F - S&F, R&T, Sdg(K), Wes. 



33 

*Emilia sonchifolia (L.) DC. 

An occasional volunteer on Palmyra. 

P - Daw. 
*Erechtites hieracifolia (L. ) Raf. 

This introduced weed was misdetermied as E. valerianaefolia (Wolf) 

DC. and appeared in the Dawson (1959) list under that name. It has 

since been redetermined by Fosberg as the closely related 

_E. hieracifolia . 

P - HiM. 
*Erigeron bonariensis L. 

=Conyza bonariensis L. 

The E_. canadensis L. which appeared in the Dawson (1959) list has 

been redetermined as E. bonariensis . This is probably the same as the 

E. albidus (WilldenowT A. Gray recorded by Christophersen (1927) on 

Fanning. 

P - HiM, Daw. 

F - H&K(UC). 
*Gaillardia pulchella Foug . 

An ornamental which has volunteered extensively around the Cable 

Station on Fanning. 

F - R&T, Wes. 
*Pluchea indica (L.) Less. 

Present on Palmyra and found in a few locations on Christmas. 

P - Bry, HiM, Daw. 

C - Ham, Sdg(K), Wes. 
*P . X fosbergii Cooperider & Galang 

This hybrid can often be found where P. indica and P. odorata grow 

together. 

P - HiM, Daw, Lng. 

C - Lng. 
*P. odorata (L.) Cass. 

= P. symphytifolia (Mill.) Gillis 

Forms dense stands on Christmas in disturbed areas. The specimen 

collected by Hill in 1949 and reported by Dawson (1959) has been 

redetermined as X P^. fosbergii . 

P - Lng. 

F - R&T, Wes 

C - Gal, Ham, Lng, Jen(K), Gri(K), Sdg(K), Wes. 
*Sonchus oleraceus L. 

Christophersen (1927) noted it around Cable Station on Fanning. 

F - [Bal] 
*Synedrella nodif lora (L. ) Gaertn. 

A common weed in disturbed areas. 

P - HiM, Daw, Lng. 

W - H&K(UC), Sdg(K), Wes. 

F - Bal, H&K(UC), R&T, Wes. 
*Tridax procumbens L. 

A local population found near Captain Cook Hotel on Christmas Island. 

May be a recent arrival. 

C - Wes. 
*Verbesina encelioides (Cav. ) B. & H. ex Gray 

A volunteer around London village on Christmas. 

C - Wes. 



34 

*Vernonia cinerea (L. ) Less. 

A common weed along roadsides and and around settlements. 

P - Daw, Lng. 

W - Ber, H&K(UC), Wes. 

F - Bal, Cht, H&K(UC), R&T, Wes. 

C - Ber, Wes. 

BIBLIOGRAPHY 

Anonymous, 1838-41, Log of the USS Peacock, Pacific Manuscripts Bureau 
Doc. 773. 

, 1874-86, Pacific Islands: Arundel, Plant Lists, Tropical Asia , 

Australasia and Southern Islands, 1874-86 , Volume 14, bound volume of 
determination lists in Royal Botanical Gardens, Kew. 

(Marie-Helene Sachet?), 1962, The Natural History Society 

(Christmas Island) and its Bulletin, Atoll Research Bulletin , 94:1-5. 

Arundel, John T. , 1870-1919, Diary , Pacific Manuscript Bureau 
manuscript nos . 480-492. 

, 1890, Phoenix Group and other islands of the Pacific. New 

Zealand Herald , 5 and 12 July, Auckland. 

Bailey, Eric, 1977, The Christmas Island story , Stacey International, 
London. 

Ball, S. C. (n.d.), Field note books , manuscript in the library of the 
Bernice P. Bishop Museum. 

Bennett, Frederick D. , 1970, Narrative of a whaling voyage round the 

globe from the year 1833 to 1836 , 2. vols., N. Israel, Amsterdam and 

Da Capo, New York (facsimile of 1840 edition published by Richard 
Bentley, London). 

Brown, F.B.H., 1930, New Polynesian plants, Occasional Papers Bernice 
P. Bishop Museum , 9:1-23. 

, 1931, Flora of southeastern Polynesia, Bulletin Bernice P. 

Bishop Museum , 84. 



Bryan, Edwin H. Jr. 1942, American Polynesia and the Hawaiian chain , 
Tongg Publishing Co., Honolulu. 



, 1974, Panala'au memoirs , Bernice P. Bishop Museum, Honolulu. 

Carter, John (Ed.), 1984, Pacific Islands yearbook , Pacific 
Publications, Sydney. 

Chock, Alvin, 1963, J. F. Rock, 1884-1962, Taxon, 12:89-102. 



35 



an d Dean c. Hamilton, 1962, Plants of Christmas Island, Atoll 

Research Bulletin , 90. 

Christophersen, Erling, 1927, Vegetation of Pacific equatorial islands 
Bulletin Bernice P. Bishop Museum , 44. 

Dawson, E. Yale, 1959, Changes in Palmyra Atoll and its vegetation 
through the agency of man, 1913-1958, Pacific Naturalist , 1:1-51. 

English, John, 1857, Letter to William Miller H.B.M. Consul General in 
the Pacific Ocean, Foreign Office Document , 58/85 Miller No. 14. 

Fosberg, F. R. 1939, Notes on Polynesian grasses , Occasional Papers 
Bernice P. Bishop Museum , 15:37-48. 



, 1943, Notes on the plants of the Pacific atolls III, a brief 

summary, Bulletin Torrey Botanical Club, 70:386-397. 



f 1953, Vegetation of Central Pacific atolls, a brief summary, 

Atoll Research Bulletin , 23. 

Gallagher, M. D. , 1960, Bird notes from Christmas Island, Pacific 
Ocean, The Ibis , 102:489-502. 

Garnett, Martin, 1981, Christmas Island wildlife sanctury: Information 
to visitors , Wildlife Conservation Unit, Christmas Island, mimeo. 

Geesink, R. , 1969, An account of the genus Portulaca Indo-Australia and 
the Pacific, Blumea , 17:275-301. 

Greene, Daniel B., 1860-65, Log of the Massachusetts, Pacific 
Manuscripts Bureau , Doc . 349 . 

Gregory, Herbert E., 1923, The report of the Director for 1923, 
Bulletin Bernice P. Bishop Museum , 4. 



, 1925, The report of the Director for 1924, Bulletin Bernice P. 

Bishop Museum , 24. 



t 1935, The report of the Director for 1934, Bulletin Bernice P. 

Bishop Museum, 133. 

Hemsley, William Botting, 1855, List of the plants collected in the 

Pacific Islands by J.T.Arundel, Report on the scientific results of 
the voyage of the H.M.S. Challenger. Botany Vol. 1, Part 4, p. 116. 

Herms, William B., 1925, Entomological observations on Fanning and 
Washington Islands, together with general biological notes, 
Pan-Pacific Entomologist, 2:49-54. 



, 1926, Diocalandra taitensis (Guerin) and other coconut pests of 

Fanning and Washington Islands, Philippine Journal Science, 
30:243-271. 



36 

Holley, Richard, 1853-57, Log of the Washington, Pacific 
Manuscripts Bureau , Doc. 369. 

Holtum, R.E., 1974, Asplenium Linn, sect. Thamnopteris, Gardener's 
Bulletin , 27:143-154. 

Jenkin, R. N. and M. A. Foale, 1968, An investigation of the coconut 
growing potential of Christmas Island , Directorate of Overseas 
Surveys, Land Resources Division (Land Resource Study, 4), Tolworth, 
England, 2 vols. 

Judd, G. P., 1859, Journal of G.P. Judd , Hawaii Mission Childrens 
Library, Honolulu. 

Hussey, Benjamin R. , 1841-45, Log of the James Maurey, Pacific 
Manuscripts Bureau , Doc. 365. 

Keyte, G.S., 1861, Fanning's Island: An incident, The Friend , 18:31. 

Kondo, Y. and Willian J. Clench, 1952, Charles Montague Cooke Jr.: A 

bio-bibliography, Special Publication, Bernice P. Bishop Museum , 42. 

Langdon, Robert, 1974, Arundel, the shy Cecil Rhodes of the Pacific 
Islands, Pacific Islands Monthly , 45:59-61. 

Lanjouw, J. and F.A. Staflen, 1954, Index Herbariorum Part II, Collectors 
A-D, Regnum Vegetabile 2, International Bureau for Plant Taxonomy and 
Nomenclature, Utrecht, Netherlands. 

Loomis, Elisha and Maria, 1820-24, Extracts from the journal of Elisha 
and Maria Loomis , typescript in Hawaii Mission Childrens Library, 
Honolulu. 

Loomis, Maria, 1820-24, Journal of Mrs. Maria Sartwell Loomis , 
typescript in Hawaii Mission Childrens Library, Honolulu. 

Luomala, Katherine, 1953, Ethnobotany of the Gilbert Islands, Bulletin 
Bernice P. Bishop Museum, 213 , 

Lucett, Edward, 1851, Rovings in the Pacific from 1837 to 1849 , 
2 vols. , Longman, Brown, Green and Longmans, London. 

Maragos, James E. , 1979, Palmyra Atoll: Preliminary environmental 
survey and assessment , U.S. Army Corps of Engineers, Honolulu. 

Martelli, V., 1926, A new species of Pa nd anus from Fanning Island, 
University of California Publications in Botany , 13:145-146. 

Merrill, E.D. , 1925, On the flora of Fanning and Washington Islands , 

Christophersen referred to this manuscript when preparing his study 
of the vegetation of the Central Pacific atolls and stated that it 
was on file in the Library of the Bernice P. Bishop Museum. No trace 
can be found of it there. 



37 

Ministry of Education Training and Culture, Kiribati Government, 1979, 
Kiribati: Aspects of history , published jointly with Pacific Studies 
and Extension Services, University of the South Pacific, Tarawa, 
Kiribati. 

Neal, Marie C. , 1965, In gardens of Hawaii , Bernice P. Bishop Museum 
Press, Honolulu. 

Poel^itz, Karl von, 1936, New species of Portulaca from southeastern 
Polynesia, Occasional Papers, Bernice P. Bishop Museum , 12:1-6. 

Republic of Kiribati, 1983, Resources Study of Fanning and Washington 
(Northern Line Islands) , mimeographed government report, 63 pp. 

Rock, Joseph F. , 1916, Palmyra Island: with a description of its flora , 
Honolulu Star Bulletin, Honolulu. 

t 1929, The voyage of the Luka to Palmyra Island, Atlantic 

Monthly , 144:360-366. 

Skerrett, Joseph S., 1873-4, Log of the U.S.S. Portsmouth , manuscript 
in the U.S. National Archives, Washington D.C. 

Sledge, W. A. 1984, University of Leeds, personal communication. 

Stone, Benjamin C. , 1968, Notes on Pandanus in the Line Islands, 
Micronesica , 4:85-93. 

St. John, Harold, 1952, A new variety of Pandanus and a new species of 
Fimbristylis from the Central Pacific Islands. Pacific Plant Studies 
11, Pacific Science , 6:145-150. 

, 1974, The vascular flora of Fanning Island, Line Islands, 

Pacific Ocean, Pacific Science , 28:339-355. 

Streets, Thomas H., 1876, Description of a new duck from Washington 
Island, Bulletin Nuttall Ornitholigical Club , 1:46-47. 

, 1977a, Contributions to the natural history of the Hawaiian 

and Fanning Islands and Lower California made in connection with the 
U.S. North Pacific Surveying Expedition 1873-75, Bulletin 
of the U.S. National Museum , 7. 

, 1877b, Some account of the natural history of the Fanning 

group of islands, American Naturalist , 11:65-72. 

Updike, John, 1973, Pigeon feathers and other stories , Knopf, New York. 

Wentworth, Chester K. 1925, A tropical peat bog, Bulletin Geological 
Society of America , 36:137. 



38 



f 1931, Geology of the Pacific equatorial islands, Occasional 

Papers Bernice P. Bishop Museum , 11. 

Wilkes, Charles, 1970, Narrative of the U. S. Exploring Expedition , 

5 vols., Gregg Press, Upper Saddle River, New Jersey, (facsimile of 
first edition published in 1845). 



ATOLL RESEARCH BULLETIN 
No. 288 



COMMUNITY STRUCTURE OF REEF-BUILDING 
CORALS IN THE FLORIDA KEYS: CARYSFORT REEF, 
KEY LARGO AND LONG KEY REEF, DRY TORT.UGAS 



BY 
Phillip Dustan 



Issued By 
THE SMITHSONIAN INSTITUTION 
Washington, D- C-, U-S-A- 
May 1985 



COMMUNITY STRUCTURE OF REEF-BUILDING 
CORALS IN THE FLORIDA KEYS: CARYSFORT REEF, 
KEY LARGO AND LONG KEY REEF, DRY TORTUGAS 

By 
Phillip Dustan 1/ 

The reefs of the Florida Keys are widely known and have drawn the 
attention of scientists since the early 1800s. The landmass of the 
Keys are the fossil remains of Pleistocene reefs (Hoffmeister and 
Multer, 1964). Their species composition can be seen in nearly every 
canal cut and rock quarry (Hodges, 1977). Receiving less attention 
however, are the reefs that make up the present living chain of reefs 
from Fowey Rocks south to the Dry Tortugas. These reefs are 
distributed in and along the outer edge of the shallow lagoon on the 
seaward side of the Keys. There are hundreds of individual reefs in 
the Keys, however there are less than 25 that could be considered to 
be more than patch reefs. The largest, most well-developed outer 
reefs presently are, in north-south order: Carysfort, Molasses, Looe 
Key, the Sambos, Long Key, and Loggerhead Reefs. Looe Key, Carysfort, 
Molasses and Long Key Reefs are similar in that they have rich coral 
communities which exhibit species zonational patterns similar to other 
Caribbean reefs (e.g. Goreau, 1959), and a topographic relief that 
appears to be the result of active coral growth on top of older reefs 
or eolian dune systems (Shinn 1963, 1977, 1980). Rates of reef 
accretion are greatest in the region of Key Largo and the Dry Tortugas 
(Shinn 1977). Although they exist at the northern and southernmost 
ends of the Florida Keys, Carysfort Reef, Key Largo and Long Key Reef 
in the Dry Tortugas are the most similar of the large reefs. Each is 
exposed to prevailing seas and has approximately the same depth range 
and species composition. 

This communication is the result of two parallel studies on the 
distribution of reef-building corals on Carysfort Reef, Key Largo and 
Long Key Reef, Dry Tortugas. The aim of the projects was to 
characterize the species composition of reef-building corals from the 
northern and southernmost localities of the Keys, establish base line 
data for future studies, and, through comparison, attempt to identify 
the impact of man on the reefs in the Key Largo area of the northern 
Florida Keys. Participants in the project include K. Lukas , J. 
Thompson, D. Girardin, K. Gordon, J. Halas, C. Richardson. Other 
contributors included J. W. Japp and J. Wheaton-Smith from the 
Department of National Resources, State of Florida, and G. Davis of 
the National Park Service. All assisted in phases of the field work 
and all are due grateful thanks. This research was supported by the 
Smithsonian Institution and Harbor Branch Foundation with logistical 
support in the Dry Tortugas provided by The National Park Service. 

1/ Department of Biology, College of Charleston, Charleston, SC 29424 



METHODS 

There are many stages of reef development and community 
complexity in both the Key Largo and Tortugas areas. The sites were 
chosen for their exposure to the prevailing seas and the general 
lushness of reef community as observed in aerial photographs and 
preliminary SCUBA excursions to each reef. As the long range goal of 
the project was to provide baseline data for both areas, the most 
information on the organization of the reef communities could be 
gathered in short time periods at the richest areas. Furthermore, 
should changes occur in the species composition of either reef, it 
might be most readily detectable in the areas of highest coral 
coverage and species diversity, as there is some suggestion that the 
most complex regions of ecosystems may be the most susceptible to 
environmental perturbation (Margalef, 1963). Carysfort Reef was 
surveyed in the spring of 1975 and Long Key in July 1975. 

The abundance and species composition of the coral community was 
estimated using line transects (Loya, 1972). This technique estimates 
projected surface-area coverage. It is biased in that flat colonies 
will project more surface area than round colonies which are spherical 
in skeleton morphology but not in tissue coverage (Porter, 1972). 
The upper surfaces of most colonies however, are covered with tissue, 
while their sides are often not. The error introduced by not using 
the chain method is probably 2-3 percent. However, the reduction in 
underwater working time afforded by the line transect allowed us to 
undertake a project of this magnitude in the relatively short period 
of time we had available. If one colony overlapped another, each 
colony was measured and recorded sequentially. This was not common in 
most parts of the reef with the main exception being the regions of 
profuse Acropora cervicornis growth on Carysfort Reef and the sides of 
the surge channels on Long Key Reef. 

The length of transects used to measure coral abundance was 
determined by running two 50 meter long transects on the fore-reef 
terrace of Carysfort Reef at a depth of 17 meters. The species that 
the line crossed were recorded for each successive meter and a species 
area curve derived. The data (Fig. 1) show that the species area 
curve reaches its asymptote in the first twenty meters. A measurement 
transect length of twenty-five meters was chosen as optimal and used 
throughout the two study areas. (For more details see Loya, 1972). 

On both reefs, measurement transects were positioned along a 
reference line set from the surface. A single long line was stretched 
from the deepest point of the reef to the shallow reef flat. 
Measurement transect lines (25m in length) were placed perpendicular 
to this line creating a grid of transects that paralleled the reef 
flat and the prevailing swell. The interval between measurement 
transects varied from 3 to 20 meters depending on the coral coverage 
and reef geomorphology . In areas of extremely steep slope the lines 
were spaced at three meter intervals, ten meters in regions of high 
coral coverage, and twenty meters apart in zones of very low coverage. 



Carysfort Reef 

Carysfort Reef (Fig. 2.) consists of two parallel platforms 
tangential to the prevailing seas. The inner platform is densely 
covered with living reef corals, while the outer platform supports a 
much reduced population of hermatypes. Seaward of the outer platform, 
fathometer recordings show small knolls between 25-35m. At forty 
meters there is a sill 3-4m high which is colonized by reef-building 
coral communities consisting mostly of Agricia spp. and Montastrea 
spp. (Dustan, Girardin, and Halas, unpublished observations). The 
angle of the slope increases seaward of this deep sill and the bottom 
drops off into the Straits of Florida. 

The study site is situated on the inner terrace of Carysfort 
Reef, approximately 75 meters south of the Carysfort lighthouse. The 
transect line runs on a compass bearing of 100 degrees magnetic from 
behind the reef flat to the edge of the first terrace at 20 meters. 
The inner terrace is approximately 250 meters wide and exhibits 
zonational patterns in the distribution of the reef-building corals 
that resemble the zonational patterns described for Jamaica (Goreau, 
1959) and the Bahamas (Storr, 1964). Zones occur in a series of 
successive bands parallel to the reef flat and thus perpendicular to 
the direction of the prevailing seas. The zonational patterns are not 
as well defined as in Jamaica and there is considerable patchiness in 
the distribution of species. The study area contains six zones based 
on changes in species composition and morphology (Fig. 3., Table 1). 
From behind the reef towards the Straits of Florida these are: 

1 . Back reef 

2. Reef flat 

3 . Acropora palmata zone 

4. Gorgonian zone 

5. Fore-reef terrace 

6 . Fore-reef escarpment 

The terminology used here is modified from Goreau (1959) and 
Kinzie (1973). This ecological zonation scheme is similar to that 
proposed for Key Largo Dry Rocks (Shinn 1963) and Grecian Rocks (Shinn 
1980). 

Carysfort Reef Zonation 

The inshore limit of Carysfort Reef, the back reef, is 
characterized by low coral coverage on a coarse sand bottom. Small 
outcrops of Montastrea annularis and Acropora cervicornis are 
colonized by encrusting Porites astreoides and Agaricia agaricites . 
The sand bottom interdigitates with the lee side of the reef flat. 
In some places there is an abrupt change between the two areas and in 
others the transition is more gradual. These ecotone areas are 
inhabited by large Mj_ annularis colonies and groves of A^. cervicornis . 
Along the irregular backside edge of the reef flat there are large 
colonies of Acropora palmata . Some colonies are overturned suggesting 
occasional heavy storm damage. 



The reef flat is approximately 50 meters wide and tabletop flat. 
It is covered with Acropora palmata and red crustose coralline algae, 
and is similar, but more expansive, than the reef flats of the inner 
reefs Grecian and Key Largo Dry Rocks (Shinn 1963, 1980) ,and those 
described for St. Croix by Adey (1975). The frame is constructed of 
densely packed dead Acropora palmata colonies in growth position which 
are encrusted with red crustose coralline algae. Other reef building 
species, mainly Porites astreoides , Agaricia agaricites , and Acropora 
cervicornis inhabit hollows that place them below the mean height of 
the reef flat. The structure of the reef flat appears dense but is 
riddled with narrow tunnels beneath the branches of the dead Acropora 
palmata framework. The seaward edge of the reef flat grades into 
irregular groves of Acropora palmata that suggest the beginnings of a 
spur and groove structure. In places the colonies are dense and 
overlap extensively, often overgrowing one another (Photo A). This 
region is the region of greatest wave activity on the reef. Close to 
the reef flat the tips of the Acropora palmata branches are level with 
the reef table and gradually deepen seaward. Further seaward, coral 
coverage decreases and the Acropora palmata colonies become oriented 
into long spurs that jut into the open sea (Shinn, 1963). Irregular 
sand channels run between the spurs with relief between the channel 
floors and top of the spurs approaching 2-3 meters in places. Coral 
rubble, mostly Acropora palmata , is strewn along the channels between 
the spurs . 

Interspersed with the Acropora palmata community are patches of 
the hydrozoan Millepora complanata in association with Porites 
astreoides , Favia fragum, the Gorgonia ventilina , and carpets of the 
zooanthid Palythoa spp . The blades of the Millepora colonies are 
oriented predominantly tangential to the prevailing seas with blades 
occasionally offset at right angles. This species association is 
analogous to the sea fan zone described by Storr (1964) for the 
Bahamas and occurs mostly on the tops of the Acropora palmata spurs 
and reef rock to a depth of approximately four meters. Millepora 
complanata is very abundant on the tops of the outcrops, comprising 
over 80% of the total coral coverage or over 45% of the reef 
substrate. This area is similar to the Millepora-Montastrea zone on 
Dry Rocks (Shinn, 1963) but more spread out and not as well organized. 

Seaward of the Acropora palmata zone there is a trough that is 
approximately 25 meters wide. The bottom consists mostly of hard reef 
rock covered with gorgonians and reef corals. Coral coverage in this 
trough drops to an estimated 20%. The morphology of the corals and 
rock coverage are similar to the shallow barren zone described by 
Kinzie (1973) for the reefs of Discovery Bay, Jamaica and appears 
analogous to Shinn's Rubble zone of the inner reefs (Shinn, 1963, 
1980). Seaward of the trough is a broken line of reef rock forming an 
irregular ridge which has a relief of 2-3 meters. This ridge 
parallels the reef crest and is dissected by numerous small channels 
and breaks. In a few places, the ridge takes on the appearance of the 
shallower spur and groove system of the Acropora palmata zone. The 
top and seaward side of the ridge supports a large sea fan community 



(Storr, 1964). Millepora complanata covers 48% of the reef substrate 
at the transect site. Other species include an occasional large 
colony of Montastrea cavernosa . M. annularis , and Colpophvllia natans . 
This ridge system terminates abruptly on the seaward side in an area 
of sparse coral coverage and the gorgonian zone begins. 

Unlike the region just described, the gorgonian zone has a sparse 
cover of small hemispherical colonies of Porites astreoides , and 
Dichocoenia stokesii and supports a rich and diverse community of 
gorgonians and algae (Photo B). The sea-fan - Mi 1 lepor a complanata 
species complex is virtually absent. The substrate is hard reef rock 
of very low relief which allows settlement of gorgonians (Kinzie, 
1973) which include members of Pterogorgia, Pseudoptergorgia and 
Eunicia. There are no surge channels or buttress features. 

Occasionally situated on this flat, gently sloping plane are large 
colonies of Montastrea annularis . In the transect area one such 
colony approached 7 meters in diameter and 3 meters in height. This 
colony sheltered a fish cleaning station and was the center of focus 
of the local fish population (Photo C). At a depth of nine meters the 
gorgonian zone terminates sharply with the sudden occurrence of 
Acropora cervicornis colonies. Coral coverage changes from less than 
10% to over 25% in less than 3 meters horizontal distance (Photo D) . 
The presence of these Acropora cervicornis marks the beginning of the 
fore-reef terrace population, an area of high species diversity and 
coverage . Just seaward of the gorgonian zone the community is 
dominated by Acropora cervicornis , Montastrea annularis , and 
Colpophvllia natans . Further seaward, the dominant species change to 
Stephanocoenia michelinii , Montastrea annularis , and Mycetophvllia 
ferox. This species assemblage in turn is replaced by a Siderastrea 
siderea dominated community on the fore-reef escarpment. Submarine 
light levels on the fore-reef terrace are relatively low due to 
turbidity (visibility is usually less than 17-18m). 

The morphology of the fore-reef slope is irregular with corals, 
gorgonians, and sponges occupying reef rock knolls separated by small 
patches of fine sediment . Most of the hermatypic corals grow upwards 
off the sediment covered bottom and then increase in surface area. 
This creates tall, slightly expanding cylindrical coral mounds 
between 0.5 and 1 meter above the soft, fine sediment covered bottom. 
In some instances the pillars are formed by a single coral colony and 
in others a few colonies, making the mound similar to a multiscoop ice 
cream cone. The sides of these coral build-ups are colonized by 
small corals, encrusting gorgonians, sponges, and bryozoa in addition 
to a rich and diverse algal community. Biological erosion appears to 
be intense and many of these pillars topple easily when jarred. These 
mounds vary in size and often coalesce when the edges of living coral 
colonies meet. This coalescence gives the reef the appearance of 
being much more solid than it really is and adds tremendously to the 
geometric complexity of the internal framework structure. Toward the 
escarpment the pillars are more isolated, rise higher off the bottom, 
and the reef framework even less solid. 



The escarpment marks the end of the fore-reef terrace. In places 
it is a vertical drop of slightly over five meters and in others a 
steep slope. Coral capped reef rocks overhang the steeper edges. 
Collapsed overhangs and slump block features are common (Photo E and 
F). In some areas large talus piles of reef rock cover the soft bottom 
at the base of the escarpment, and in other places the reef gradually 
grades into a soft, fine sediment substrate. On a large scale the 
escarpment appears to be irregularly buttressed. The buttresses seem 
to be constructed by reef-building corals growing on old slump blocks 
and reef debris. These buttress features are approximately 30 meters 
apart. Seaward of the escarpment is a soft fine sediment covered 
bottom that stretches flat some 100 meters to the beginning of the 
outer terrace. 

The outer terrace supports a sparse coral population similar in 
species composition and morphology to the gorgonian zone. This 
assemblage is characteristic of the outer slopes of Molasses, French, 
and Elbow reefs in Key Largo. Large colonies of Montastrea annularis 
are scattered sporadically over the bottom and, as in the gorgonian 
zone, support diverse fish populations and frequently, cleaner fish 
stations. There are small sand channels running seaward and most 
species of coral are usually small. 

Long Key Reef 

Long Key Reef lies at the southeastern edge of the Dry Tortugas 
platform. The study site lies southeast from Fort Jefferson .facing 
southeast, the direction of the prevailing swell, and away from the 
direction of most winter storms. The prevailing current pattern over 
the Tortugas platform is from northwest to southeast such that the 
water passing over the reef drains from across the entire reef 
platform (Davis, 1982). This green to blue-green water is laden with 
organic debris and fine sediment. Estimated horizontal visibility was 
almost always less than 10 meters during our field session earning the 
nickname "shadowland" for the reef. The reef supports a large 
diverse fish population (Jones and Thompson, 1975), along with 
associated reef algae and gorgonians. Long Key was chosen as the 
study site as it is the only reef in the Dry Tortugas that displays 
zonational patterns and geomorphology similar to Carysfort Reef, our 
primary work site in the northern Keys. 

The morphology of Long Key Reef may be seen in Fig. 4, a 
fathometer tracing which was run over the transect site on a compass 
heading of 120 degrees magnetic. The reef is backed by a shallow 
lagoon which leads into a reef flat composed of coral rubble. The 
seaward edge of the reef flat slopes very gently to a depth of 
approximately 10 meters where the slope increases. Seaward of the 
reef are a few scattered gorgonian and coral encrusted rocks. The 
reef measures just slightly under four hundred meters from the flat to 
the base of the reef at 18-20m. Long Key Reef (Fig. 5, Table 2.) may 
be divided into five distinct parallel zones which lie parallel to the 



reef flat and tangent to the prevailing seas: 

1 . Lagoon 

2. Reef flat 

3 . Patch reef zone 

4. Gorgonian zone 

5. Spur and groove 

The lagoon was not surveyed with transects. It consists of small 
patches of Acropora cervicornis . and Porites porites . There is one 
small grove of Acropora palmata situated in a channel. This is the 
only living stand of this species in the area to our knowledge, 
although the species was much more abundant in the 1800's. In 1976 a 
cold water thermal shock killed two-thirds of the small stand (Davis, 
1982) 

The reef flat is composed of loose coral rubble, mostly Acropora 
cervicornis , Porites spp . and a few Acropora palmata fragments. At 
high tide it is submerged and is frequently exposed at low tide. The 
flat appears to be the result of the accumulation of debris tossed up 
by storms and not the end product of in-situ of coral growth. The 
gently sloping shallows in front of the flat are sparsely covered with 
Porites porites , Porites astreoides , Siderastrea radians and support a 
dense, fleshy algal population. This region stretches approximately 
fifty meters seaward. 

Patch reefs appear when the water depth approachs 5 meters. The 
algal coverage decreases and the bottom is covered with coarse 
carbonate sediment, mostly shelly sands and coral fragments. Dotted 
on this are small rock islands of reef rock that stand 30-50 cm off 
the bottom. These islands support a dense gorgonian population and 
about ten species of coral (Photo G) . Situated among the reef rock 
islands are small stands of Acropora cervicornis ranging in size from 
0.5-2 meters in diameter. Conspicuously absent are any living 
colonies of Acropora palmata . An extensive search did turn up a small 
grove consisting of two or three dead colonies. The encrustation and 
erosion of their surfaces suggested they died between two and ten 
years previously. The patch reef region extends for approximately 
one hundred and thirty meters seaward and ranges in depth from 3-7 
meters. 

Seaward of the rock islands is an area characterized by soft 
sediment which supports a very few species of coral at low densities 
and a luxuriant population of the gorgonian Pseudopterogorgia bipenata 
(Photo H) . This region extends approximately forty to fifty meters. 

The coral population becomes more abundant and diverse as the 
spur and groove region is approached. Species diversity and coverage 
increase sharply and the reef begins to take on the appearance of a 
"true" coral reef. The spur and groove region consists of long spurs 
of coral 2-4 meters in height off the bottom which are 3-15 meters in 
width and are oriented (120 degrees magnetic) into the prevailing 



8 

swell. The principal reef building coral species is Montastrea 
annularis , which appears in a variety of growth forms from knobby 
multilobate to large flow sheets of the skirted ecotype (terminology 
after Dustan, 1975). Colonies approach four meters in diameter and 
appear to be responsible for the construction of the reef spurs (Photo 
I). The floors of the grooves are sediment covered with occasional 
pieces of loose coral rubble. In a few locations bare reef rock was 
observed in the grooves, probably the result of scouring by the 
prevailing swell. This region extends for approximately one hundred 
meters and ends at the base of the reef. It is the most diverse and 
richly populated region. The colony size of most species reaches a 
maximum in this region as well. 

The internal structure of the spurs between 10 and 18 meters is 
honeycombed with caves due to the profuse overlapping of coral 
colonies. In many instances it appears that one large coral colony 
develops into a mushroom shaped structure creating a cave beneath it. 
The floors of these caves are covered with fine sediment. The walls 
are covered with sponges and bryozoans. Fluorescene dye was released 
into these crevices in an attempt to determine the extent of the 
labyrinth. Dye released into holes would flow out others 5 to 10m 
away suggesting that the reef structure is open beneath the veneer of 
living coral. Large coral colonies of a variety of species appear 
sporadically along the seaward edge of the escarpment. Such species 
are Madracis decactis (Photo J), Agaricia lamarcki, Stephanocoenia 
michelinii , Montastrea cavernosa , and Montastrea annularis . Of these, 
only Montastrea annularis is commonly larger than a meter in diameter 
in other habitats on the reef. 

Discussion of Carysfort Reef 

Carysfort Reef marks the northern extent of lush populations of 
reef-building corals along the eastern coast of North America. Mayor 
(1914) suggested that reduced water temperature further north limited 
the northern extent of reef development. South of Key Largo reef 
development may be limited by tidal passes that allow water from 
Florida Bay to flow onto the shelf platform on the ebb tide (Ginsburg 
and Shinn 1964) or, conversely, allow cool subsurface water from the 
Straits of Florida to intrude into the shallows of the shelf platform 
(Dustan, et. al_. , 1976). Thus Carysfort appears to be situated at or 
just south of the thermal tolerance point of active reef development 
and paradoxically, is one of the most well developed reefs in the 
entire Florida Keys. 

The vertical distributions of coral coverage and number of 
species are almost independent of each other on the reef flat and in 
the shallows. Coral coverage and the number of species become more 
closely correlated in deeper water (Fig. 3). Maximal species number 
occurs at the outer edges of the terraces, seaward of maximal coral 
coverage, hinting that spatial competition in these areas may be 
intense, or that subtle environmental differences between a terrace 
and break in slope may reorder community structure (Porter, 1972). 



Substrate heterogeneity at the edge of a break in slope, or water 
circulation enhancement may create a more favorable environment for 
larval settlement as well. In any case, it must be remembered that 
reef growth, with the exception of the reef flat, is proceeding 
upwards so that today's edge is part of tomorrow's terrace. As such 
the species composition of the reef may be influenced by the 
morphology of the reef which, in turn, affects the future species 
composition. Active reef growth in shallow water on Carysfort is a 
function of the growth of Acropora palmata and Millepora complanata . 
The seaward geomorphology of the reef flat and the spur and groove 
formations are formed mostly by Acropora palmata as first described 
for Key Largo Dry Rocks by Shinn (1963). Deeper, it appears that 
Montastrea annularis and other massive corals contribute to the growth 
of the reef frame. 

The species composition of the gorgonian zone on Carysfort is 
similar to the outer reef slopes of other reefs in the northern Keys: 
Molasses, Elbow, French Reefs. On these reefs the gorgonian 
population gradually decreases as water depth increases towards the 
Straits of Florida. On Carysfort, however, this zone ends abruptly at 
10 meters where it is replaced by a diverse coral community. The 
appearance of a rich coral community and the disappearance of the rich 
gorgonian community at 10m occurs as a sharp line and is very apparent 
even to a casual visitor to the reef. This suggests that some 
environmental parameter change sharply at this depth, strongly 
influencing reef community development (Photo D) . 

Waves heading into Carysfort Reef first meet with the outer 
terrace some 300 meters seaward of the reef (Fig. 2). The shallowest 
depth of the platform is 10 meters so that wave energy below 10 meters 
is attenuated . The outer terrace thus shields the fore-reef terrace 
population from the full force of the prevailing swells and storm 
seas that pound the other outer reefs. We have witnessed this 
sheltering phenomena while diving on Carysfort during seas of 1-2 
meter wave height. At depths below 10 meters the surge is greatly 
reduced and occurs as a sharp boundary just at the beginning of the 
fore reef coral population. In the winter and spring months we have 
observed sharp discontinuities in temperature and underwater 
visibility between 10 meters and the top of the escarpment at 15 
meters. Whenever these differences in water masses have been observed 
warmer, clearer water overlays cooler, more turbid water. Either the 
cold water is intruding into the surface waters along the edge of the 
Gulf Stream as noted in the Dry Tortugas (Dustan et. al ., 1976) or 
cooler water from the back reef area is becoming trapped in the moat 
region between the inner and outer platform of Carysfort. The 
reduction of wave action results in finer sediments on the fore-reef 
slope and escarpment at Carysfort than on neighboring reefs. 
Sediments from these zones on Carysfort have sponge boring chips, 
spicules and fine sedimentary particles in great abundance. The 
sediments at comparable depths on Molasses Reef consist of much 
coarser carbonate particles. It is conceivable that the moat region 
of Carysfort serves as a sink for the deposition of fine sediment and 



10 

therefore may contain a detailed record of the depositional history of 
the northern Florida Keys. 

Discussion of Long Key Reef 

The species zonation on Long Key Reef follows the classical 
pattern first described for West Indian coral reefs (Goreau, 1959), 
but there are some distinct differences. Most conspicuous of these 
are the absence of Acropora palmata and an associated reef flat 
community. The reef flat appears to be formed by the accumulation of 
coral skeleton rubble rather than infilled, cemented Acropora palmata 
skeletons in growth position as seen at Carysfort and so 
characteristic of other Florida Keys Reefs (Shinn, 1963, 1977). 
Seaward of the reef flat where one would expect to find groves of A. 
palmata the substrate is covered mostly with algae and gorgonians, and 
a few scattered corals. Observations along the edge of the reef flat 
subsequently turned up a stand of dead A^ palmata (in addition to the 
one known living stand mentioned earlier) in growth position 
suggesting that while this species may inhabit the zone it suffers 
high mortality and never reaches the population densities found 
elsewhere in the Florida Keys. Reports of a massive coral mortality 
as a result of "black water" are mentioned in the first Carnegie 
Reports of the Dry Tortugas (Mayer, 1902), and in 1977 a cold water 
(13 degrees C) intrusion resulted in the death of almost all the 
Acropora cervicornis on the platform (Davis, per. comm) . Thus 
periodic climactic fluctuations, possibly combined with severe 
storms, may prevent Acropora palmata from establishing itself as a 
major reef-building species in the sediment laden water of the Dry 
Tortugas. 

Along with the noted absence of Acropora palmata is the absence 
of a Millepora-gorgonian species association commonly seen in the Keys 
and Bahamas. Both species occur in the Dry Tortugas but do not form 
the assemblage so common elsewhere. The assemblage is commonly found 
on the tops of shallow reef flat spurs constructed by Acropora 
palmata . Possibly the absence of Acropora palmata and the habitat its 
structure creates results in the deletion of the Millepora-gorgonian 
species complex. Conversely, the lack of a similar necessary 
ecological condition (high surf, clear water, favorable temperature) 
may be the controlling factor in the distribution of both assemblages. 



Coral coverage is closely correlated with species number on Long 
Key Reef (Fig. 5). The absence of high coverage in shallow water is 
attributed to an absence of Acropora palmata and its associated 
Millepora-gorgonian species complex. Coverage is highest deeper than 
10m where the most active reef accretion appears to be occurring. 



11 

CONCLUSIONS 

The species composition and zonation patterns of a coral reef 
are the result of species' differential abilities to settle, adapt, 
and survive the prevailing environmental conditions. The 
environmental parameters of light, water temperature and wave action, 
sedimentation, and food availability all have been thought to be of 
primary importance to corals. Biological interactions between and 
among species operate at organizational levels within this adaptive 
framework (Porter, 1974; Glynn, 1976; Connell, 1978). The interplay 
of biological and physical factors result in higher order interactions 
that determine community structure (Futuyma, 1979). Coral communities 
at both study sites show a positive correlation between average colony 
size and percentage cover. On Long Key Reef increases in coral 
coverage are the result of all species becoming more abundant. On 
Carysfort Reef increases in cover are sometimes the result of single 
species dominance, as in the Acropora palmata zone, or a general 
increase in all species as seen on the fore-reef terrace. 

The differences in patterns suggest that different environmental 
and biological pressures control the development of these two reef 
communities. Part of the reason for the differences in these two 
reefs may lie in their positions relative to the path of the Gulf 
Stream. Carysfort Reef lies at the edge of the Gulf Stream and is 
often bathed in its waters, while the Dry Tortugas are approximately 
10-20 miles north of the edge of the Gulf Stream and only occasionally 
experience clear oceanic water. The prevailing patterns of water 
movement over the Dry Tortugas result in a northwest to southeast flow 
so that the water passing over Long Key Reef has drained from the Dry 
Tortugas Platform. Thus there appears to be major differences in the 
quality of the water over the two reefs. In addition, the Gulf Stream 
buffers the population at Carysfort against cold water intrusions in 
the winter so that even though it is much farther north than Long Key, 
the minimum water temperatures are somewhat higher. There are reports 
(Mayer, 1914) of elevated water temperatures occurring in the Dry 
Tortugas in the summer coincident with periods of calm and low spring 
tides which resulted in the death of corals. Temperatures in shallow 
parts of the reef ranged from 33-38 degrees C. On Carysfort Reef this 
type of localized water heating seems unlikely as the reef is situated 
far from any other large geomorphic structures and the water is kept 
moving longshore by the Gulf Stream. 

Both reefs (Long Key and Carysfort) face into the direction of 
the prevailing winds and swell. However, Carysfort does not receive 
the full force of the swell at depths below 10 meters as the fore-reef 
terrace is protected by the second platform seaward of it. This 
platform blocks the deeper swell and may help to explain the absence 
of well defined surge channels on Carysfort Reef. Observations on the 
species composition of the seaward slope of the outer platform have 
shown it to be similar to the outer slopes of neighboring reefs that 



12 

are not protected from the swell. These regions are sparsely covered 
by coral and give the appearance of being "wave battered". The rich 
and diverse coral population on the fore-reef terrace of Carysfort 
Reef below 10 meters is not typical of other reefs in the Keys. 
Whether or not this atypical species assemblage is the direct result 
of a decrease in wave shock, change in food supply or sediment 
composition, or some other factor cannot be determined at this time; 
but remains as an intriguing question to be attacked at a future 
date . 

One of the initial objectives of this research was to study 
differences between the two reefs in an attempt to dissect out the 
impact of man on Carysfort Reef. However, upon completion of the Long 
Key Reef survey it became apparent that the two reefs are very 
different in structure and form as a result of a suite of different 
environment parameters. There are some general observations that 
deserve comment however, and though they have not been quantified, 
they may be instructive as to the mechanisms behind man's impact on 
coral reefs. 

The incidence of broken coral colonies is high on Carysfort Reef. 
Broken colonies must expend metabolic energy to their wounds and 
regrow skeleton lost to damage. It is hard to single out the greatest 
cause of physical damage to reefs by man but it is fair to say that 
constantly occurring damage as a result of anchoring, diving, and 
fishing is slowly but surely decreasing the amount of framework 
carbonate that corals add to the reef structures of the northern keys, 
and it is apparent on Carysfort Reef. Commensurate with these 
observations is the high incidence of corals with broken or damaged 
tissue as a result of excess sedimentation, algal overgrowth and algal 
disease (Dustan, 1977). Again, such mortality factors were 
occasionally seen in the Dry Tortugas, most notable is the havoc 
caused by anchoring in the lee of Loggerhead Key (Photo K, Davis, 
1977). 

The comparison presented in this study has provided an initial 
look at the community structure of two reefs at the opposite ends of 
the Florida Keys. Differences in community structure appear to be the 
result of local environmental differences and local geomorpho logical 
features. Severe periodic environmental perturbations may control the 
distribution of less tolerant species ( Acropora palmata and Acropora 
cervicornis ) in the Dry Tortugas. The Gulf Stream may reduce the 
probability of similar perturbations occurring on Carysfort Reef. Both 
reefs may be exposed to storms which may also affect their species 
composition and geomorpho logy . Resurveys of these communities in the 
future will begin to reveal their temporal as well as spatial 
variability. 



13 
References 

Adey, W.H. (1975) Algal ridges and coral reefs of St. Croix, their 
structure and Holocene development. Atoll Res. Bull. 187:1-66 

Connell, J.H. (1978), Diversity in tropical rain forests and 
coral reefs, Science 199(4335), 1302-1310. 

Davis, G.E. (1977), Anchor damage to a coral reef on the coast of 
Florida, Biol. Conserv. 11:29-34. 

Dustan, P. (1975), Genecological differentiation in the reef-building 
coral Montastrea annularis , Ph.D. thesis, State 
University of New York at Stony Brook. 

Dustan, P. (1975a), Growth and form in the reef-building coral 
Montastrea annularis , Marine Biology 33:101-107. 

Dustan, P., W. Japp and J. Halas (1976), Notes on the 

distribution of members of the Class Sclerospongiae, Lethaia 9(4), 
419-420. 

Dustan, P. (1977), Vitality of reef coral populations off Key Largo, 
Florida: recruitment and mortality, Environmental Geology 2:51-58. 

Futuyma, D. (1979), Evolutionary Biology . Sinaur Associates 
(Sunderland, Mass.), 565 pp. 

Ginsburg, R.N. and E.A. Shinn (1964), Distribution of the reef- 
building community in Florida and the Bahamas ( abs), Amer. 
Assoc. Pet. Geol. Bull 48:527. 

Glynn, P.W. (1976), Some physical and biological determinants 
of coral community structure in the Eastern Pacific, Ecol. 
Monographs 46(4), pp. 431-456. 

Goreau, T.F. (1959), The ecology of Jamaican coral reefs I, 
species composition and zonation, Ecology 40:67-90. 

Goreau, T.F. and J.W. Wells (1967), The shallow-water Scleractinia of 
Jamaica: revised list of species and their vertical distribution 
range, Bull. Mar. Sci. 17:442-453. 

Hodges, L.T. 1977. Coral size and orientation relationships of the Key 
Largo limestone, FLorida. Proc. 3rd Int. Coral Reef Symp. (Univ. 
Miami, Fl). pp. 348-352 

Hoffmeister, J.E. and H.G. Multer (1964), Geology and origin of the 
Florida Keys, Geol. Soc. Amer. Bull. 75: 1487-1502. 



14 

Jones, R.S. and M. J. Thompson (1978), Comparison of Florida reef fish 
assemblages using a rapid visual technique, Bull. Mar. Sci. 
28(1): 159-172. 

Kinzie, R.A.K. Ill (1973), The zonation of West Indian gorgonians, 
Bull. Mar. Sci. 23:93-155. 

Loya, Y. (1972), Community structure and species diversity of 

hermatypic corals at Eilat, Red Sea, Mar. Biol. 13:100-123. 

Mayer, A.G. (1902). The Tortugas, Florida as a station for research 
biology. Science 17:190-192 

Mayor, A.G. (1914), The Effects of Temperature on Tropical Marine 
Animals, Carnegie Institute, Washington, Pub. 183, 6:1-24, 

Porter, J.W. (1972), Patterns of species diversity in Caribbean reef 
corals, Ecology 53:745-48. 

Porter, J.W. (1974), Community structure of Coral Reefs on Opposite 
Sides of the Isthmus of Panama, Science 186:543-545. 

Shinn, E.A. (1963), Spur and groove formation on the Florida Reef 
Tract. Jour. Sed. Pet. 33,2: 291-303. 

Shinn, E.A. (1980), Geologic history of Grecian Rocks, Key Largo 
Coral Reef Marine Sanctuary. Bull. Mar. Sci. 30,3: 646-656. 

Shinn, E.A., J. H. Hudson, R.B.Halley, and B. Lidz (1977), Topographic 
control and accumulation rate of some Holocene coral reefs: 
South Florida and Dry Tortugas. Proc. 3rd Int. Coral Reef Symp . 
Vol 2: Geology. (Univ. Miami, Fl.) pp. 1-7 

Storr, J.F. (1964), Ecology and Oceanography of the Coral-Reef Tract, 
Abaco Island, Bahamas. GSA Special Papers No. 79. 



15 



Table 1: Coral species coverage on different morphological zones 
Carysf ort Reef , Key Largo 



Species 



RF 



Mean Percent Coverage by Zone 
A. pal Trough Ridge GZ FRT 



FRE 



Acropora palmata .... 
Acropora cervicornis . . 
Mycetophyllia lamarckana 
Mycetophyllia ferox . . 
Mycetophyllia danana . . 
Mycetophyllia aliciae 
Solenastrea hyades . . . 
Agaricia agaricites . . 
Agaricia lamarcki . . . 
Agaricia fragilis . . . 
Helioseris cucullata . . 
Colpophyllia natans . . 
C. breviserialis .... 
Scolymia cubensis . . . 
Scolymia lacera .... 

Mussa angulosa 

Montastrea annularis . . 
Montastrea cavernosa . . 
Manicina areolata . . . 

Favia fragum 

Siderastrea radians 
Siderastrea sidera . . 
Dichocoenia stokesii . 
Stephanocoenia michelinii 
Diploria strigosa . . . - 
Diploria clivosa .... - 
Diploria labyrinthiformis - 
Isophyllia sinuosa ... 
Isophyllastraea rigida . - 
Porites porites . . . 
Porites astreoides . . 
Porites furcata . . . 

Madracis spp 

Madracis decactis . . 
Millepora alcicornis . 
Millepora complanata . 
Eusmilia fastigiata 

Number of species 

in each zone 6 

Number of transects 



Distance along transect 
from base of reef (m) 






36.2 


9.0 




















3.6 








30.4 

















0.2 


<1 


<1 














0.2 


1.6 


<1 




















<1 




















1.2 


<1 


1.0 


1.7 


2.9 


<1 


<1 


1.0 

















<1 




















<1 











<1 


<1 





<1 




















2.2 


<1 

















7.6 


<1 

















<1 


<1 


<1 











<1 


17.5 


1.7 











1.1 





<1 


1.1 





<1 


<1 





<1 




















<1 











<1 


<1 





<1 


3.3 


6.8 














<1 


<1 

















<1 


1.5 


2.2 



<1 













<1 


<1 


<1 


.7 


1.5 


3.4 


4 


.2 


<1 


<1 


1 



















<1 






















<1 






















<1 


<1 








<1 







<1 


<1 





.4 


7.2 


10.5 


48 


.0 





7.0 


<1 



















<1 


<1 


6 


7 


9 




5 


11 


24 


17 


4 


3 


3 




1 


2 


5 


3 


320- 


250- 


200- 




140 


100- 


80- 


30 


260 


230 


160 






120 


40 






Note: RF=back reef + reef flat , A.pal= Acropora palmata zone, 
GZ= gorgonian zone, FRT=f ore-reef terrace, FRE= fore reef escarpment, 
Dash signifies presence on Long Key but absent on Carysfort Reef. 



16 



Table 2: Coral species coverage on different morphological zones 
Long Key Reef, Dry Tortugas 

Species Mean Percent Coverage by Zone 

Patch Gorgonian Spur and Groove 
Reefs 1 2 3 



Acropora palmata 

Acropora cervicornis <1 

Mycetophyllia lamarckana ... 

Mycetophyllia ferox 

Mycetophyllia danana 

Mycetophyllia aliciae .... 

Solenastrea hyades - 

Agaricia agaricites <.5 

Agaricia lamarcki 

Agaricia fragilis 

Helioseris cucullata 

Colpophyllia natans 

C. breviserialis 

Scolymia cubensis 

Scolymia lacera 

Mussa angulosa 

Montastrea annularis 

Montastrea cavernosa 

Manicina areolata 

Favia fragum 

Siderastrea radians <.5 

Siderastrea sidera <.5 

Dichocoenia stokesii 

Stephanocoenia michelinii . . 

Diploria strigosa 

Diploria clivosa 1.5 

Diploria labyrinthiformis . . 

Isophyllia sinuosa 

Isophyllastraea rigida . . . .<.5 

Porites porites <.5 

Porites astreoides <1 

Porites furcata 

Madracis spp - 

Madracis decactis 

Millepora alcicornis <.5 

Millepora complanata 

Eusmilia fastigiata 

Number of species 

in each zone 9 

Number of transects 4 



<1 


<1 


4.6 








<1 


<1 


<1 








2.2 


<1 








<1 











<1 





<.5 


<1 


1 


<1 











5.2 








<1 


<1 








<1 


<1 





<1 


2 


<1 








<1 














<1 











<1 








<1 








<1 


15.0 


10.3 


<1 


<1 


4.6 


6.8 





<1 




















<.5 


4.5 


4.8 


7.5 


<.5 


<1 











<1 


1.2 


3.0 


<.5 











<1 





<.5 





<1 


<.5 


<1 





<.5 


<.5 


<.5 











<.5 





1.2 


9 


1 


<1 


1.1 


1 


2.1 


1 








<1 


<1 


<1 


<1 


<1 


<1 



12 
8 



<.5 



16 
6 



<.5 



24 

5 



17 

5 



Distance along transect 
from base of reef (m) 



370- 
300 



290- 
180 



160- 
110 



100- 
50 



40- 




Note: Dash (-) signifies presence on Carysfort but absent on Long Key 



17 

Table 3. Species List for Areas of Study 

Species Dry Long Key Key Carysfort 

Tortugas Reef Largo Reef 

Acropora palmata * - * * 

Acropora cervicornis * * * * 

Mycetophyllia lamarckana ... * * * * 

Mycetophyllia ferox * * * * 

Mycetophyllia danana * * * * 

Mycetophyllia aliciae .... * * * * 

Solenastrea hyades * * * *1 

Agaricia agaricites * * * 

Agaricia lamarcki * * * * 

Agaricia fragilis * * * * 

Helioseris cucullata * * * * 

Colpophyllia natans * * * * 

Colpophyllia breviserialis . . * * * 

Scolymia cubensis * * * 

Scolymia lacera * * * * 

Mussa angulosa * * * * 

Montastrea annularis * * * 

Montastrea cavernosa * * * * 

Dendrogyra cylindrus 

Manicina areolata * * 

Favia fragum * * * 

Favia conferta * 

Siderastrea radians * * * * 

Siderastrea sidera * * * 

Dichocoenia stokesii * * * 

Dichocoenia stellaris .... * * * * 

Stephanocoenia michelinii . . * * * * 

Diploria strigosa * 

Diploria clivosa _ _ * 

Diploria labyrinthif ormis . . * - * * 

Isophyllia sinuosa * * * * 

Isophyllastraea rigida .... * * * 

Porites porites * * * * 

Porites astreoides * * * * 

Porites furcata - - * * 

Madracis spp * * * * 

Madracis decactis * * * * 

Millipora alcicornis * * * * 

Millipora complanata * - * 

Eusmilia fastigiata * * * * 

Oculina diffusa * * * 

Cladocora spp * 

Meandrina meandrites * * * * 

* = present, - = absent 
1: outer terrace only 




METER NUMBER 



Fig. 



1. Species-area curve for reef coral species on Carysfort Reef, 
Key Largo, Florida. Transects were run at a depth of 17 
meters on the fore-reef terrace of the study site. 




300 250 200 100 
TRANSECT METER NUMBER 



Fig. 2. Fathometer tracing of Carysfort Reef showing the overall 
morphology of the reef. Note the terrace 6eaward of the 
study site which rises to a depth of 10 meters. 



70-1 



z 

8 Z H 

z u, = 9 < - 

< o § ' a s! 2 

z — o 9 u. ^ 

S " H l g < 

O T 

a 




r-25 



-20 « 



-15 



-10 



HORIZONTAL DISTANCE (Meters) 
CARYSFORT REEF TRANSECT METER NUMBER 



Fig. 3. Graph depicting the percentage coral cover, number of 
species, and depth profile of Carysfort Reef study site. 

Or- ^ 



25 






X 
H 

W 
Q 



50- 



75 



100 



— ' — » ^^ 
















LONG KEY REEF 












DRY TORTUGAS, FLA. 












'iMj 1 1 
jMj approx. 100m 


- 


-J 
< 








ROOVE 

1 




O 


tu 


Z 


O UJ 




H 
-< 

tu 


<Z 

>>o 
ac n 
to 


UJ 

= o 

UN 


< 

g UJ 


Z ts) 

oi 


Q Z 






< 


pi ™ 

o 


D 


5< 

< j 


p£ 


tL. 


a. 


o 




t/3 


CO D- 



350 250 200 150 100 

TRANSECT METER NUMBER 



Fig. 4. Fathometer tracing of Long Key Reef, Dry Tortugas showing 

the overall morphology of the reef. Note the absence of an 
offshore terrace. 



-J 




< 


< 

-J 


< z 


U. 


5- O 


U. 


X 55 


UJ 


2 


OC 






u. 








r 25 i 


1 


~ 


\ 


-20 2 






V) 


-is y 




EL. 


s 


V5 

-10 £ 


EC 




(— 


* ■* 


cu 


-5 uj 


UJ 


a 


Q 


-o i 






z. 



HORIZONTAL DISTANCE (Meters) 
LONG KEY REEF TRANSECT METER NUMBER 



Fig. 5. Graph depicting the percentage coral cover, number of 

species, and depth profile of Long Key Reef study site. 



Photograph Legends 

A. Dense thickets of Acropora palmata on the seaward edge of the 
reef flat of Carysfort Reef. 

B. A rich gorgonian, sponge and algal community inhabits the 
gorgonian zone on the gorgonian zone on the fore-reef terrace 
from about 5m to 10m. Scleractinian corals such as 

P. Porites (2) and D. stokesii (1) comprise 

less than 10% of the total coverage. The colony of 

D. stokesii is approximately 12cm in diameter. 

C. Large colonies of M. annularis occur sporadically in the gorgonian 
zone of Carysfort reef. Such colonies act as islands on the plain 
and become centers of focus for fish and invertebrate populations. 
This colony is approximately 3m high and 6.4m in greatest diameter. 

D. The ecotone between the gorgonian zone (right) and the fore- 
reef slope (left) is extremely sharp (arrows). Coral coverage 
changes from less than 10% to greater than 25% in less than 3 
meters horizontal distance at a depth of 10m on Carysfort Reef. 

E. A large spreading colony of Agaricia spp. is covering a large 
block of reef rock, apparently a talus block from an earlier 
s lump . 

F. View of the escarpment of Carysfort Reef, 18m. Virtually every 
colony in the view is surrounded by fine sediment which collects 
on the escarpment as a result of wave attenuation by the outer 
reef platform. 

G. A reef rock island on the fore-reef terrace of Long Key Reef. 
This particular island has been formed by flU. annularis and 
M.alcicornis . 5m 

H. Luxuriant population of P_j_ bipenata inhabits a 50m wide 

zone on Long Key Reef at a depth of 4-6 meters . The largest of 
these colonies are 1.5 to 2m in height. 

I. A narrow sand channel slowly being overgrown by a large colony of 
M. annularis illustrates the hollowness of the spur and 
groove zone on Long Key Reef. 12m. 

J. Diver Karen Lukas examining an exceptionally large colony of 
M. mirabilis at the base of Long Key Reef, 18m. 

K. An anchor, probably lost by a fishing boat, embedded in a patch of 
A. cervicornis on Long Key Reef. Note the broken and dead 
rubble surrounding the shank. The coral is regenerating and will 
eventually overgrow the anchor, 10m. 




A. Dense thickets of Acropora palmata on the seaward edge of the 
reef flat of Carysfort Reef. 




B. A rich gorgonian, sponge and algal community inhabits the 
gorgonian zone on the gorgonian zone on the fore-reef terrace 
from about 5m to 10m. Scleractinian corals such as 
P. Porites (2) and D. stokesii (1) comprise 
less than 10% of the total coverage. The colony of 
D. stokesii is approximately 12cm in diameter. 




C. Large colonies of M. annularis occur sporadically in the gorgonian 
zone of Carysfort reef. Such colonies act as islands on the plain 
and become centers of focus for fish and invertebrate populations . 
This colony is approximately 3m high and 6.4m in greatest diameter. 




D. The ecotone between the gorgonian zone (right) and the fore- 
reef slope (left) is extremely sharp (arrows). Coral coverage 
changes from less than 10% to greater than 25% in less than 3 
meters horizontal distance at a depth of 10m on Carysfort Reef. 




A large spreading colony of Agaricia spp. is covering a large 
block of reef rock, apparently a talus block from an earlier 
s lump . 




I 




F. View of the escarpment of Carysfort Reef, 18m. Virtually every 
colony in the view is surrounded by fine sediment which collects 
on the escarpment as a result of wave attenuation by the outer 
reef platform. 




G. A reef rock island on the fore-reef terrace of Long Key Reef , 
This particular island has been formed by M^ annularis and 
M.alcicornis . 5m 




H. Luxuriant population of P^ bipenata inhabits a 50m wide 

zone on Long Key Reef at a depth of 4-6 meters . The largest of 
these colonies are 1.5 to 2m in height. 




I. A narrow sand channel slowly being overgrown by a large colony of 
M. annularis illustrates the hollowness of the spur and 
groove zone on Long Key Reef. 12m. 




J. Diver Karen Lukas examining an exceptionally large colony of 
M. mirabilis at the base of Long Key Reef, 18m. 




K. An anchor, probably lost by a fishing boat, embedded in a patch of 
A. cervicornis on Long Key Reef. Note the broken and dead 
rubble surrounding the shank. The coral is regenerating and will 
eventually overgrow the anchor, 10m. 






ATOLL RESEARCH BULLETIl 
No- 289 



THE DISTRIBUTION, ABUNDANCE AND PRIMARY PRODUCTIVITY 
OF SUBMERGED MACROPHYTES IN A BELIZE BARRIER-REEF 
MANGROVE SYSTEM 



By 
Iark M- Littler, Phillip R. Taylor, 
Diane S. Littler, Robert H. Sims 
and James N. Norris 



Issued By 

THE SMITHSONIAN INSTITUTION 

Washington D- C, U.S-A- 

May 1985 



100 m 




SOUTH POINT 



Figure 1. Location of Twin Cays study sites on the Belize barrier 
reef. 



THE DISTRIBUTION, ABUNDANCE AND PRIMARY PRODUCTIVITY 
OF SUBMERGED MACROPHYTES IN A BELIZE BARRIER-REEF 
MANGROVE SYSTEM 

By 
Mark M. Littler* Phillip R. Taylor** 
Diane S- Littler*. Robert H. Sims* 
and James N. Norris' 



* 



ABSTRACT 

The comparison of wave-exposed (bay) to sheltered (channel) 
macrophyte assemblages in a Belize mangrove system revealed higher 
standing stocks of productive filamentous algae in the latter, 
correlated with relatively low levels of physical disturbance from sea 
urchin herbivory and wave turbulence. The sheltered channel site, 
while containing fewer total species and lower species richness, 
exceeded the bay site in total cover and species evenness. The 
Shannon-Weaver index of diversity was nearly equal at both sites. 
Five species comprised 96% of the cover at the bay site, led by* the 
jointed calcareous alga Halimeda opuntia f. triloba (37%) and 
Thalassia testudinum (26%); whereas, H. opuntia f. triloba (40%), 
Amphiroa fragilissima (22%) and J_« testudinum (16%) provided the 
majority of the total community productivity. At the channel site, 
six taxa contributed 96% of the cover, dominated by a mat-forming, 
gelatinous, filamentous species of naviculoid diatom (29%) and 
Caulerpa verticillata (28%). Major primary producers at the channel 
site were the three cover dominants, the gelatinous diatom (24% of the 
total community carbon fixed), C_. verticillata (22%) and H. opuntia f. 
triloba (20%). The total daylight community primary productivities at 
the two sites (bay = 17.2, channel = 13.4 grams carbon fixed per meter 
squared of substratum per day ranked among the higher rates recorded 
for dense seagrass beds and were considerably higher than those 
reported for most calcareous reef flat habitats. This high apparent 
photosynthetic potential may be related to reduced levels of herbivory 
and a greater availability of recycled nutrients near mangrove 
islands. 



Department of Botany, National Museum of Natural History, 
Smithsonian Institution, Washington, D.C. 20560 

icic 

Present address: Biological Oceanography Program, National 
Science Foundation, Washington, D.C. 20550 



INTRODUCTION 

The fringing-reef margins of tropical atolls and coastal zones 
represent shallow subtidal to intertidal calcareous frameworks with 
diverse epibiota that have received intensive study in recent years. 
Diverse algal standing stocks have been described for Curacao, Dutch 
West Indies (e.g. Van Den Hoek et al. 1975) and Saint Croix, U.S. 
Virgin Islands (Connor and Adey 1977) in the Caribbean. Pacific 
reefs, such as Enewetak Atoll, U.S. Trust Territory (Odum and Odum 
1955); Waikiki reef, Hawaii (Doty 1971; Littler 1973a); Guam, U.S. 
Trust Territory (Tsuda 1971); American Samoa (Dahl 1972); Kaneohe Bay, 
Hawaii (Smith 1973); and Heron Island, Australia (Hatcher 1982; 
Hatcher and Larkum 1983), have been the subjects of comparable 
studies . 

While mangrove communities coincide with the worldwide 
distributions of calcareous biotic reefs and dominate many of the 
world's tropical and subtropical coastal zones, relatively little work 
has been done on the standing stocks of benthic macrophytes on the 
flats that border mangrove islands. On mangrove islands, like coral 
islands, organisms comprise the major structural elements, but unlike 
most biotic reef communities, mangrove islands are intertidal, endure 
wider fluctuations in temperature and salinity and tend to contain 
silty submerged substrata. Because of their interesting 
characteristics and a paucity of background information, we initiated 
a quantitative survey of mangrove macrophyte distributions, abundances 
and productivities as a necessary basis for the design of further, 
more specialized experimental studies (e.g. Taylor, Littler and 
Littler, submitted). 

A classification of marine plant functional-form groups [see 
Littler, Littler and Taylor (1983) for definitions] has been used to 
interpret (1) productivity patterns over broad geographic areas 
(Littler and Arnold 1982), (2) evolutionary changes with respect to 
fluctuations in herbivory through geological time (Steneck and Watling 
1982), (3) holistic views of tropical barrier-reef seaweed ecology 
(Littler, Littler and Taylor 1983), (4) biogeographical responses of 
algae to herbivory (Gaines and Lubchenco 1982) and (5) the effects of 
disturbance on subtropical (Littler and Littler 1984) and temperate 
(Murray and Littler 1984) macrophyte communities. The functional- 
group approach is an effective mechanism for assessing complex 
community patterns without having to tediously deal with the 
demography of each of the component species. Consequently, we felt it 
would be instructive to analyze the Twin Cays macrophyte populations 
from this framework. 



METHODS AND MATERIALS 

Study Areas 

_ Twin Cays is a mangrove system (16 50"N, 80 06'W) representative 
of similar pristine offshore islands within the lagoonal portion of 
the Belize barrier system. The island (Fig. 1; see also Rlltzler and 
Macintyre 1982) is divided by a main channel that grades toward each 
opening into shallow beds of Thalassia testudinum Banks ex KHnig. 

After extensive reconnaissance of the Twin Cays shoreline, the 
precise location of the upper end of each study transect was 
determined (by consensus of several experienced marine ecologists) 
along biologically representative portions of two major benthic 
habitat types. The channel study site, located on the east side of 
the main channel, is protected from wave action, with only moderate 
tidal currents controlling water exchange. The area studied is a 
typical subtidal mud bank extending 5.5 m outward to a dropoff from 
intertidal Rhizophora mangle Linnaeus into the meandering 0.5 to 4.0 m 
deep main channel dominated by Thalassia and Caulerpales. In 
contrast, the less-sheltered bay site on the northern margin of the 
island (Fig. 1) typifies habitats where waves and currents are 
greater, frequently resulting in the mud banks being partially eroded 
to form vertical walls or undercut ledges. The mud bank studied is 
6.5 m wide and terminates abruptly where an undercut bank extends down 
to a 3.0 m deep silt substratum with sparsely scattered plants of the 
algal genus Caulerpa [C_. mexicana (Sonder) KUtzing and £. 
sertularioides (Gmelin) Howe]. 

Transects 

Data were obtained on 11-12 April 1980 by photographing numbered 
quadrats perpendicular to the substratum with a 35 mm Nikonos camera 
equipped with an electronic flash unit and using Kodachrome 64 
transparenty film. Each quadrat contained a plastic label affixed to 
the upper left corner that was marked with a wax pencil to identify 
permanently each of the photosamples. 

In the laboratory, the developed transparencies were projected 
onto a sheet (21 x 28 cm) of white bristol paper. The paper contained 
a grid pattern of dots at 2.0 cm intervals on the side of the 
transmitted light; this has been shown (Littler and Murray, 1975) to 
be an appropriate density (i.e. 1.0 per cm ) for consistently 
reproducible estimates of cover. The number of dots superimposed on 
each species was then scored twice (i.e. replicated after movement of 
the grid) with the percentage cover values expressed as the number of 
"hits" for each species divided by the total number of dots contained 
in the quadrats. Reproducibility was high and seldom varied more than 
+ 5% for a given species. Species that were not abundant enough to be 
scored by the replicated grid of point intercepts were assigned a 
cover value of 0.1%. In cases of multi-layered communities, more than 
one photograph per quadrat was taken to quantify each stratum after 
upper strata had successively been moved aside, often yielding total 
biotic coverages of greater than 100%. 



The method as applied here does not allow for the quantification 
of microalgae (small epiflora or inflora) when they occur in low 
abundances. We realize that these may be metabolically very active, 
but their analysis requires special techniques and expertise, which 
comprise separate problems in themselves. For this reason, our 
measurements were restricted to macrophytes that could be discerned in 
the field with the unaided eye. However, we did quantify microflora 
(e.g. mats of a filamentous diatom) when it occurred in high 
abundances. Twenty three contiguous quadrats along a 6.65 m transect 
were taken at the bay site while 31 quadrats along a 7.25 m line were 
sampled at the channel site. 

Community Productivity 

Net apparent photosynthesis of the most abundant macrophytes 
found at the study sites was determined in a shallow current channel 
at ambient water temperatures (27 C) on 24 April 1980. Four replicate 
incubations per taxon were conducted between 0900 and 1430 hrs under a 
photon flux of 900 to 1900 micro Einsteins/m /sec of 

photosynthetically active radiation (45,000 to 95,000 lux). This was 
the natural light in situ and within the range of light saturation 
values documented for other shallow macroalgal species (King and 
Schramm 1976; Arnold and Murray 1980; LaPointe et al. 1983). Net 
productivity was measured to 0.1 parts per million of dissolved oxygen 
by means of YSI Model 57 oxygen analyzer and calculated as milligrams 
carbon fixed per unit of thallus area per hour assuming a 
photosynthetic quotient of 1.00. To enable comparisons with other 
tropical marine ecosystems, daily daylight rates were approximated by 
multiplying the mean hourly rates by the number of daylight hours 
above the saturation intensity. All specimens used were from shallow 
locations in full sunlight. The methods concerning the selection of 
material, handling, incubation and oxygen analysis were within the 
limits recommended by Littler (1979) and Littler and Arnold (1980). 

Analyses of Data 

Data obtained by photogrammetric sampling enable quantification 
of the distributions and abundances of standing stocks in relation to 
transect distance and depth. All quadrat data were summed and 
averaged to yield mean cover values and used to interpret differences 
in macrophyte populations and communities between sites. The two- 
dimensional (i.e. planar area intercepting the light) species cover 
per square meter of substratum in conjunction with individual 
productivities per square meter of planar thallus area were multiplied 
to estimate the contribution of each abundant macrophyte to overall 
community production. 

Diversity measurements have been widely employed by those 
responsible for assessing the effects of disturbances on biotic 
communities. Species diversity is often measured by indices (see 
Poole 1974 or Pielou 1975 for references and definitions) that include 
components of both species richness and equitability (the evenness 
with which the individuals are apportioned among species. The problem 
with any single index is that both the richness and equitability 
components of diversity are confounded. Many diversity indices also 






contain the underlying assumption that the ecological importance of a 
given species is proportional to its abundance. We avoided these 
problems by using the commonly-applied Shannon and Weaver H' index 
(incorporating both richness and evenness) along with separate indices 
for richness (counts of taxa, Simpson's Index, Margalef's D') and 
evenness [equitability (E'), Pielou's J']. These were calculated for 
the cover data using natural logarithms of macrophyte cover and used 
as supplementary information to provide between-site comparisons of 
community structure. 

To characterize natural species assemblages within each site 
grouping in an unbiased manner, the cover data for all quadrats were 
subjected to hierarchical cluster analyses (flexible sorting; Smith 
1976) using the Bray and Curtis (1957) percentage distance statistic. 
This produced dendrograms of transect assemblages that were then 
interpreted according to their dominant biota and environmental 
affinities and used to map the prevalent zonal patterns for the two 
sites. 



RESULTS 

The channel site exceeded the bay site in all measured parameters 
(cover, productivity and diversity; Table 1) except species richness. 
Parameters for which the channel site was higher were as follows: 
total cover (1.5 times), total benthic community primary productivity 
(1.3X), Pielou's evenness index (J', 1.2X) and equitability (E', 
1.5X). Conversely, Margalef's richness index (D') was 1.8 times 
higher at the bay site, while Simpson's index was 1.2 times greater. 
Shannon-Weaver diversity values (H') for the two sites were nearly 
equal, since the higher richness at the bay site balanced the greater 
evenness at the channel area. 

Five species comprised 96% of the cover at the bay site (Table 
2), led by Halimeda opuntia f. triloba (Decaisne) Barton (37%) and 
Thalassia testudinum Banks and Konig (26%); whereas, H. opuntia f. 
triloba (40%), Amphiroa fragilissima (Linnaeus) Lamouroux (22%) and T. 
testudinum (16%) provided the majority of the total community 
productivity. At the channel site (Table 3), six taxa contributed 96% 
of the cover, dominated by a filamentous, gelatinous diatom species 
(29%) and Caulerpa verticillata J. Agardh (28%). Major primary 
producers at the channel site were the three cover dominants, the 
gelatinous diatom (24% of the total community carbon fixed), £. 
verticillata (22%) and H. opuntia f. triloba (20%). 

The cluster plots (Figs. 2 and 3) reveal distinctly different 
zonational patterns between the two sites. The bay site is 
characterized by (1) the Caulerpa / Halimeda assemblage forming a band 
nearest the mangrove roots which intergrades into (2) a zone dominated 
by the Halimeda / Dictyota assemblage. This last cluster group, in 
conjunction with (3) an assemblage characterized by the corallines 
Amphiroa and Neogoniolithon , forms a third zone, while the fourth 
most-seaward zone is delineated by the dominance of (4) a 
Thalassia / Halimeda assemblage. The channel site also contained four 
zonal assemblages as follows, proceeding from the edge of the mangrove 



island toward the main channel: (1) a gelatinous diatom-dominated 
group overlying C. verticillata in very silty substrata nearest the 
mangrove roots, with Halimeda also present more channelward, (2) a 
Spyridia / Halimeda assemblage, followed by (3) a zone dominated by the 
Halimeda spp. cluster and, lastly, (4) a Thalassia assemblage that 
continues on across the Twin Cays main channel (Fig. 1). 

In terms of production rates per square meter of thallus planar 
area (Fig. 4), the corallines Neogoniolithon strictum (Foslie) 
Setchell and Mason (0.38 g carbon/m /h) Amphiroa fragilissima (0.29) 
and Amphiroa rigida Lamouroux var. antillana BjSrgesen (0.29) were 
highest, followed by the calcareous green alga Penicillus pyriformis 
A. & E.S. Gepp (0.19). In terms of production per unit of substratum, 
the channel site was nearly 1.3 times more productive on the average 
than the bay site, led by the gelatinous diatom (4.1 g carbon/m of 
substratum/day). In contrast, the siphonaceous green alga, Halimeda 
opuntia f. triloba was by far the greatest contributor to community 
productivity (5.4 g C/m /d) at the bay site. 



DISCUSSION 

Although the seagrass Thalassia testudinum was a conspicuous 
component of the floras at both sites, algae collectively were the 
predominant organisms (cf. Tables 2 and 3), providing 73.8% of the 
plant cover and 83.6% of the productivity at the bay site and 89.9% 
and 92.4% of the total plant cover and productivity at the channel 
site, respectively. The filamentous-group dominated production at the 
channel site; whereas, the soft-bottom siphonaceous algal forms 
provided the majority of production at the bay site. The overall 
community contribution of the seagrass was more than double in terms 
of both cover (26.2% vs. 10.1%) and productivity (16.4% vs 7.6%) at 
the bay site. 

McRoy and Lloyd (1981) have contrasted marine macrophytes into 
two fundamentally different groups: (1) the macroalgae and (2) the 
seagrasses. The former group, according to these authors, is 
characterized as analogous to filter feeding animals (in their 
extraction of nutrients) while secured to two-dimensional hard 
substrata by means of a holdfast. The latter extract nutrients from 
both the water column and soft sedimentary, three-dimensional 
substrata by means of vascular root-rhizome systems that also serve to 
anchor them. This dichotomy ignores the many siphonaceous algal forms 
that we have shown (Tables 2 and 3) to be prevalent in association 
with Twin Cays mangrove islands. These algae, mainly of the order 
Caulerpales, also have root-like and rhizomatous systems for 
attachment in soft substrata and, because cross walls are minimal, can 
utilize cytoplasmic streaming to translocate nutrients taken up from 
both the sedimentary and aquatic milieu. 

In terms of the predominant marine plant functional groups (Table 
4), the Jointed-Calcareous-Group dominated the bay site with 48.6% of 
the total cover, followed by the Thick Leathery-Group (26.0%), Sheet- 
Group (12.6%), Filamentous-Group (11.7%), Crustose-Group (1.2%) and 
Coarsely-Branched-Group (0.1%). The Jointed-Calcareous-Group also 



contributed a disproportionately large amount to the total marine 
macrophytic productivity (63.1%), the order of importance to 
production of the remaining groups was the same as for cover (Thick- 
Leathery, 16.1%; Sheet, 9.7%; Filamentous, 7.5%; Crustose, 3.4%; 
Coarsely-Branched, 0.1%). Members of the Sheet-Group and the 
Crustose-Group were largely absent from the channel site. For the 
channel site, the Filamentous-Group contributed the majority of total 
community cover (64.9%) as well as productivity (58.6%), whereas the 
Jointed-Calcareous-Group ranked second (20.8% of total cover and 29.8% 
of total productivity) and the Thick-Leathery-Group was third (10.2% 
of cover and 7.6% of productivity, respectively). Coarsely-Branched- 
and Crustose-Groups were minor components at both sites. 

The two macrophyte communities are essentially quite similar 
(Tables 2 & 3), with the majority of differences (e.g. cover, 
productivity, evenness and richness; Table 1) due to the epiphytic 
overstory contributed by the Filamentous-Group at the sheltered 
channel site. This extensive mat-like canopy of a very delicate, 
filamentous/gelatinous diatom and weakly anchored Caulerpa 
verticillata may be very susceptible to wave damage at the more 
northerly, exposed bay site. We observed mats of the diatom being 
torn loose by the action of a boat wake on one occasion. The poor 
biomechanical resistance of filamentous algae has been documented for 
both temperate habitats (Littler and Littler 1980) and biotic reefs 
(Littler, Littler and Taylor 1983). 

Mangrove island mud banks, such as those studied here, tend to be 
depauperate in regard to the photosynthetic corals, the Sheet-Group of 
macrophytes, non-calcified frondose forms (i.e. Coarsely-Branched- and 
Thick-Leathery-Group) and the Crustose-Group. In contrast, on hard- 
surfaced carbonate reefs (Littler and Littler 1984), corals, non- 
articulated coralline algae (Crustose-Group) and/or various small 
microscopic forms (Filamentous-Group) usually comprise the major 
cover. Larger non-calcareous frondose macrophytes (Sheet-, 
Coarsely-branched-, Thick-leathery-Groups) occur abundantly on 
reef flats (Doty 1971; Wanders 1976; Connor and Adey 1977), 
unstructured sand plains (Earle 1972; Dahl 1973; Hay 1981a) or 
deep-water sites (Littler et al. 1985) where herbivory is very 
low. 

The inconspicuousness of non-calcified algae on many shallow reef- 
front systems is thought (Randall 1961; Wanders 1977; Borowitzka 1981) to 
result primarily from intensive grazing by the numerous herbivores and 
omnivores inhabiting these spatially heterogeneous systems. Where spatial 
heterogeniety (i.e. protective cover for fishes and sea urchins) is 
minimal on tropical reefs, herbivore activity is relatively low (Connor 
and Adey 1977; Brock 1979; Hay et al. 1983) and reasonably large standing 
stocks of macrophytes often develop (Doty 1971; Tsuda 1971; Connor and 
Adey 1977; Wanders 1976). On Twin Cays, the shallow bordering mud flats 
are extremely low in spatial heterogeneity, with the macrophytes 
themselves comprising most of the three-dimensional structure. Barracuda 
(Sphyraenidae), mangrove snapper (Lut janidae), jacks (Carangidae) and 
other fishes are abundant predators near the channels and hanging roots of 
mangrove islands (personal observations), and this undoubtedly contributes 
to the reduced levels of herbivorous fishes. 



Sea urchins can be locally abundant within the mangrove root 
habitat and often produce a grazing halo (cf. Ogden et al. 1973) that 
tends to be dominated by grazer-resistant (Paul and Fenical 1983) 
species of Halimeda (Figs. 4 & 5) adjacent to and between the 
Rhizophora mangle . Thalassia testudinum with other interspersed 
seagrasses and Caulerpales become abundant beyond the feeding ranges 
of urchins unless an eroded bank prevents such a transition. Sea 
urchins were numerous among the mangrove roots at the bay site and 
moved onto the algal/seagrass flats at night. Total sea urchin 
density in the vicinity of the bay site was 4.4 *m , comprised mostly 
of Echinometra viridis Agassiz, E. lacunter (L.) and Lytechinus 
variegatus (Lamarck), with Diadema antillarum Philippi, Eucidaris 
tribuloides (Lamarck) and Tripneustes ventricosa (Lamarck) also 
present. During approximately 20 person-hours of searching, no 
urchins were encountered in the vicinity of the channel site. Other 
research (Taylor, Littler and Littler, submitted) indicates that such 
differences in herbivory may, in conjunction with the physical action 
of waves mentioned above, account for the reduction of delicate 
filamentous forms and epiphytes and the predominance of herbivore- 
resistant macrophytes at the bay site. 

Such resistant plant populations often contribute (Rogers and 
Salesky 1981) a major portion of the total primary productivity of 
some reefs. However, most evidence (e.g. Wanders and Wanders-Faber 
1974, Bunt 1975, Marsh 1976, Dahl 1976 Larkum 1981, Rogers and Salesky 
1981) indicates that it is the fast-growing and opportunistic 
filamentous algae of sparse mats that result in the very high 
primary production rates per unit area of most biotic reefs. 
Conversely, tightly-compacted mats of algae (turfs), such as 
those of the channel site, usually show reduced productivity 
levels (Littler and Arnold 1980, Hay 1981b, Taylor and Hay 1984) 
due to overlapping diffusion gradients and self shading. 

The total community primary productivity at the channel site was 
28% higher than at the bay site (17.2 vs. 13.4 g C fixed'm -2 of 
substratum'd ). This difference was largely due to the contributions 
of the filamentous/gelatinous diatom (4.1 g C fixed'm" *d ) and 
Spyridia f ilamentosa (2.2), epiphytes that were not abundant at the 
bay site. 

The macrophytic daylight community productivities at both sites 
were quite high relative to reef systems and compare favorably with 
the upper rates reported from dense seagrass meadows (Table 5). 
Reported daily primary productivities of seagrass communities span the 
upper range from 5.8 to 18.7 g carbon/m 2 /d. Rates reported for reef 
systems range upwards to 7.2 g C/m /d. We conclude that mud reef- 
flats adjacent to the mangrove islands of the Belize barrier reef 
system produce at rates comparable to dense seagrass beds and are 
considerably more productive than typical carbonate reef-flat 
habitats. This high photosynthetic capacity may be related to 
reductions in herbivory, enabling larger standing stocks to develop, 
and the recycling of nutrients from decompositional processes, which 
would be expected to augment the primary productivity of these 
otherwise nutrient-impoverished waters. 



ACKNOWLEDGEMENTS 

We appreciate the support made available as part of the 
Smithsonian Institution's Western Atlantic Mangrove Project, ably 
directed by K. Rlltzler. Additional sponsorship was provided by S. 
Dillon Ripley through the Secretary's Fluid Research Fund. This paper 
is Contribution No. 153 of the Smithsonian Institution's Reef and 
Mangrove Study, partly funded by the Exxon Corporation. 



REFERENCES 

Arnold, K. E. and S. N. Murray. 1980. Relationships between 

irradiance and photosynthesis for marine benthic green algae 
(Chlorophyta) of differing morphologies. J. Exp. Mar. Biol. 
Ecol. 43: 183-192. 

Bakus, G. J. 1967. The feeding habits of fishes and primary 
production at Eniwetok, Marshall Islands. Micronesica 3: 
135-149. 

Borowitzka, M. A. 1981. Algae and grazing in coral reef ecosystems. 
Endeavour, 5: 99-106. 

Brawley, S. H. and W. H. Adey. 1977. Territorial behavior of 

threespot damself ish ( Eupomacentrus planifrons ) increases reef 
algal biomass and productivity. Envir. Biol. Fish. 2: 45-51. 

Bray, J. R. and J. T. Curtis. 1957. An ordination of the upland 

forest communities of southern Wisconsin. Ecol. Monogr. 27: 325- 
349. 

Brock, R. E. 1979. An experimental study on the effects of grazing 
by parrotfishes and role of refuges in benthic community 
structure. Mar. Biol. 51: 381-388. 

Buesa, R. J. 1972. Production primaria de las praderas de Thalassia 
testudinum de la platforma noroccidental de Cuba. INP, Cuba 
Cent. Inv. Pesqueras. Reva. Bal. Trab. CIP, 3: 101-143. 

Bunt, J. S. 1975. Primary productivity of marine ecosystems. In: H. 
Lieth and R. H. Whittaker (Eds.), Primary Productivity of the 
Biosphere. Ecological Studies, Springer-Verlag, Berlin 14: 169- 
183. 

Connor, J. L. and W. H. Adey. 1977. The benthic algal composition, 
standing crop, and productivity of a Caribbean algal ridge. 
Atoll. Res. Bull. 211: 1-15. 

Dahl, A. L. 1972. Ecology and community structure of some tropical 
reef algae in Samoa. Pages 36-39, in K. Nisizawa (Ed.), 
Proceedings of the Seventh International Seaweed Symposium. 
University of Tokyo Press, Tokyo, Japan. 



10 

Dahl, A. L. 1973. Surface area in ecological analysis: 

quantification of benthic coral-reef algae. Mar. Biol., 23: 
239-249. 

Dahl, A. L. 1976. Generation of photosynthetic surface area by coral 
reef algae. Micronesica 12: 43-47. 

Doty, M. S. 1971. Physical factors in the production of tropical 

benthic marine algae. Pages 99-121 rn J. D. Costlow, Jr., (Ed.) 
Fertility of the Sea. Gordon and Breach, New York. 

Earle, S. A. 1972. The influence of herbivores on the marine plants 
of Great Lameshur Bay, with an annotated list of plants. Pages 
17-44 in B. B. Collette and S. A. Earle (Eds.), Results of the 
Tektite Program: Ecology of Coral Reef Fishes, Science Bulletin 
14, Los Angeles County Natural History Museum, Los Angeles, 
California. 

Gaines, S. D. and J. Lubchenco. 1982. A unified approach to marine 
plant-herbivore interactions. II. Biogeography. Ann. Rev. Ecol. 
Syst. 13: 111-138. 

Gessner, F. and L. Hammer. 1960. Die primarproduktion in 

Mediterranean Caulerpa - Cymodocea Wiesen. Bot. Mar. 2: 157-163. 

Hatcher, B. G. 1982. The interaction between grazing organisms and 
the epilithic algal community of a coral reef: a quantitative 
assessment. Pages 515-524 _rn E. D. Gomez et al. (Eds.), 
Proceedings, Fourth International Coral Reef Symposium, Marine 
Science Center, Quezon City, Philippines 

Hatcher, B. G. and A. W. D. Larkum. 1983. An experimental analysis 
of factors controlling the standing crop of the epilithic algal 
community on a coral reef. J. Exp. Mar. Biol. Ecol. 69: 61-84. 

Hay, M. E. 1981a. Herbivory, algal distribution, and the maintenance 
of between-habitat diversity on a tropical fringing reef. Am. 
Nat. 118: 520-540. 

Hay, M. E. 1981b. The functional morphology of turf-forming 

seaweeds: persistence in stressful marine habitats. Ecology 62: 
739-750. 

Hay, M. E., T. Colburn and D. Downing. 1983. Spatial and temporal 
patterns in herbivory on a Caribbean fringing reef: the effects 
on plant distribution. Oecologia (Berlin) 58: 299-308. 

Johnston, C. S. 1969. The ecological distribution and primary 

production of macrophytic marine algae in the eastern canaries. 
Int. Revue ges. Hydrobiol. 54: 473-490. 

King, R. J. and W. Schramm. 1976. Photosynthetic rates of benthic 
marine algae in relation to light intensity and seasonal 
variations. Mar. Biol. 37: 215-222. 



11 

LaPointe, B. E., K. R. Tenore and C. J. Dawes. 1983. Interactions 
between light intensity and temperature on the physiological 
ecology of Gracilaria tikvahiae (Gigartinales, Rhodophyta). I. 
Growth, photosynthesis, and respiration. Mar. Biol. 

Larkum, A. W. D. 1981. Marine primary productivity. Pages 370-385 
in M. N. Clayton and R. J. King (Eds.), Marine Botany, an 
Australasian Perspective. Longman Cheshire, Melbourne. 

Littler, M. M. 1973a: The population and community structure of 
Hawaiian fringing-reef crustose Corallinaceae (Rhodophyta, 
Cryptonemiales). J. Exp. Mar. Biol. Ecol. 11: 103-120. 

Littler, M. M. 1973b. The productivity of Hawaiian fringing-reef 
crustose Corallinaceae and an experimental evaluation of 
production methodology. Limnol. Oceanogr. 18: 946-952. 

Littler, M. M. 1979. The effects of bottle volume, thallus weight, 
oxygen saturation levels, and water movement on apparent 
photosynthetic rates in marine algae. Aquat. Bot. 
7: 21-34. 

Littler, M. M. and K. E. Arnold. 1980. Sources of variability 

in macroalgal primary productivity: sampling and interpretative 
problems. Aquat. Bot. 8: 141-156. 

Littler, M. M. and K. E. Arnold. 1982. Primary productivity of 

marine macroalgal functional-form groups from southwestern North 
America. J. Phycol. 18: 307-311. 

Littler, M. M. and D. S. Littler. 1980. The evolution of thallus 

form and survival strategies in benthic marine macroalgae: field 
and laboratory tests of a functional form model. Am. Nat. 116: 
25-44. 

Littler, M. M. and D. L. Littler. 1984. Biogenesis of tropical 
reefs: the contribution of algae. In: F. E. Round and D. J. 
Chapman (Eds.), Progress in Phycological Research, 3: (in press). 

Littler, M. M., D. S. Littler, S. M. Blair and J. N. Norris. 

1985. Deepest known plant life discovered on an uncharted 
seamount. Science 227: 57-59. 

Littler, M. M., D. S. Littler and P. R. Taylor. 1983. Evolutionary 
strategies in a tropical barrier reef system: functional- 
form groups of marine macroalgae. J. Phycol. 19: 229-237. 

Littler, M.M. and S.N. Murray. 1975. Impact of sewage on the 
distribution, abundance and community structure of rocky 
intertidal macro-organisms. Mar. Biol. 30: 277-291. 

Marsh, J. A., Jr. 1976. Energetic role of algae in reef ecosystems. 
Micronesica 12: 13-21. 



12 

McRoy, C. P. and D. S. Lloyd. 1981. Comparative function and 

stability of macrophyte-based ecosystems. Pages 473-489 .in A. R. 
Longhurst (Ed.), Analysis of marine ecosystems. Academic Press, 
New York. 

Murray, S. N. and M. M. Littler. 1984. Analysis of seaweed 

communities in a disturbed rocky intertidal environment near 
Whites Point, Los Angeles, Calif. U.S.A. In: Proceedings, 
Eleventh International Seaweed Symposium, Qingdao, China. 
Hydrobiologia, (in press). 

Odum, H. T. 1963. Productivity measurements in Texas Turtle grass 

and the effects of dredging an intracoastal channel. Publ. Inst. 
Mar. Sci. Texas 9: 45-58. 

Odum, H. T. and C. M. Hoskin. 1958. Comparative studies of the 

metabolism of marine waters. Publ. Inst. Mar. Sci. Univ. Texas 
5: 16-46. 

Odum, H. T. and E. P. Odum. 1955. Trophic structure and productivity 
of a windward coral reef community on Eniwetok Atoll. Ecol. 
Monogr. 25: 291-320. 

Ogden, J. C, R. A. Brown and N. Selesky. 1973. Grazing by the 

echinoid Diadema antillarum Philippi: formation of halos around 
West Indian patch reefs. Science 182: 715-717. 

Paul, V. J. and W. Fenical. 1983. Isolation of halimedatrial: 

chemical defense adaptation in the calcareous reef-building alga 
Halimeda . Science 221: 747-749. 

Pielou, E. 1975. Ecological diversity. John Wiley and Sons, New 
York, New York. 

Poole, R. W. 1974. An introduction to quantitative ecology. 
McGraw-Hill, New York, New York. 

Qasim, S. Z. and P. M. A. Bhattathiri. 1971. Primary production of a 
seagrass bed on Kavaratti Atoll (Laccadives). Hydrobiologia 38: 
29-38. 

Randall, J. E. 1961. Overgrazing of algae by herbivorous marine 
fishes. Ecology 42: 812. 

Rlltzler, R. and I. G. Maclntyre (Eds.) 1982. The Atlantic Barrier 
Reef Ecosystem at Carrie Bow Cay, Belize, I. Structure and 
Communities. Smithson. Contrib. Mar. Sci., No. 12. Smithsonian 
Institution Press, Washington, D.C. 

Rogers, G. S. and N. H. Salesky. 1981. Productivity of Acropora 

palmata (Lamarck), macroscopic algae, and algal turf from Tague 
Bay Reef, St. Croix, U.S. Virgin Islands. J. Exp. Mar. Biol. 
Ecol. 49: 179-187. 



13 

Smith, R. W. 1976. Numerical Analysis of Ecological Survey Data. 
Ph.D. dissertation, University of Southern California, Los 
Angeles . 

Smith, S. V. 1973. Carbon dioxide dynamics: a record of organic 

carbon production, respiration, and calcification in the Eniwetok 
reef flat community. Limnol. Oceanogr. 18: 106-120. 

Sournia, A. 1976. Oxygen metabolism of a fringing reef in French 
Polynesia. HelgolHnder wiss. Meeresunters. 28: 401-410. 

Steneck, R. S. and L. Watling. 1982. Feeding capabilities and 

limitation of herbivorous molluscs: a functional group approach. 
Mar. Biol. 68: 299-319. 

Taylor, P. R. and M. E. Hay (in press). The functional morphology of 
intertidal seaweeds: the adaptive significance of aggregate vs. 
solitary forms. Mar. Ecol. Prog. Ser. 

Taylor, P. R., M. M. Littler and D. S. Littler (in review). 
Coexistence and non-coexistence escapes of herbivory as 
structuring forces in mangrove island macroalgal communities. 
Ecology. 

Tsuda, R. T. 1971. Morphological, zonational, and seasonal studies 

of two species of Sargassum on the reefs of Guam. Pages 40-44 .in 
K. Nisizawa (Ed.), Proceedings, Seventh International Seaweed 
Symposium. Wiley & Sons, New York. 

Van Den Hoek, C, A. M. Cortel-Breeman and J. B. W. Wanders. 1975. 
Algal zonation in the fringing coral reef of Curacao, Netherlands 
Antilles, in relation to zonation of corals and gorgonians. 
Aquat. Bot. 1: 269-308. 

Wanders, J. B. W. 1976. The role of benthic algae in the shallow 

reef of Curacao (Netherlands Antilles). I: primary production in 
the coral reef. Aquat. Bot. 2: 235-270. 

Wanders, J. B. W. 1977. The role of benthic algae in the shallow 

reef of Curacao (Netherlands Antilles) III: the significance of 
grazing. Aquat. Bot., 3: 357-390. 

Wanders, J. B. W. and A. Wanders-Faber. 1974. Primary productivity 
of the S.W. coast shallow reef and the N.E. coast brown algae of 
Curacao. Br. Phycol. J. 9: 223-224. 



14 

Table 1. Indices of richness, evenness and diversity, based on 

macrophytic species numbers and cover, for the bay and channel 
sites. 



Indices Bay Channel 



H' index 2.32 2.39 

J' evenness 0.85 1.00 

E' equitability 0.64 0.99 

D' richness 2.52 1.44 

Number of species 14 11 

Simpsons index 0.11 0.09 

Total cover 97.8 151.8 

Total productivity 13.4 17.2 



(g carbon/m^/d) 



Table 2. The major contributors of community cover and primary 
productivity (g carbon/m^/d) at the bay site. 



Taxa Cover Productivity 

(%) (m of substratum) 

Halimeda opuntia f. triloba (Decaisne) Barton 

Thalassia testudinum Banks & Konig 

Dictyota dichotoma (Hudson) Lamouroux 

Caulerpa verticillata J. Agardh 

Amphiroa f ragilissima (Linnaeus) Lamouroux 

Neogoniolithon strictum (Foslie) Setchell & Mason 

Halimeda monile (Ellis & Solander) Lamouroux 

Dasya rigidula (KUtz.) Ardiss. 

Amphiroa rigida Lamouroux var. antillana B<5rgensen 

Valonia ventricosa J. Agardh 

Wrangelia sp. 

Hypnea sp. 

Penicillus capitatus Lamarck 

Dictyosphaeria cavernosa (Forssk.) B^rgesen 

Totals 97.8 13.4 



36.1 


5.4 


25.6 


2.2 


12.4 


1.3 


11.2 


1.0 


10.0 


2.9 


1.2 


0.5 


0.4 


0.07 


0.4 


— 


0.3 


0.0 


0.07 


— 


0.04 


— 


0.02 


— 


0.02 


0.01 


0.02 


0.01 



15 



Table 3. The major contributors of community cover and primary 
productivity (g carbon/m /d) at the channel site. 



Taxa 



Cover Productivity 
(%) (m of substratum) 



Gelatinous diatom 

Caulerpa verticillata J. Agardh 

Halimeda opuntia f. triloba (Decaisne) Barton 

Thalassia testudinum Banks & Ktinig 

Spyridia f ilamentosa (Wulfen) Harvey 

Halimeda monile (Ellis & Solander) Lamouroux 

Penicillus capitatus Lamarck 

Dictyosphaeria cavernosa (Forssk.) B^rgesen 

Caulerpa mexicana (Sound.) J. Agardh 

Amphiroa fragilissima (Linnaeus) Lamouroux 

Penicillus pyriformis A. & E.S. Gepp 

Totals 



43.6 


4.1 


41.9 


3.8 


22.6 


3.4 


15.4 


1.3 


13.2 


2.2 


9.0 


1.5 


2.8 


0.3 


2.7 


0.3 


0.5 


0.04 


0.08 


0.2 


0.08 


0.02 



151.8 



17.2 



Table 4. Functional-group categories and major component taxa. 



Functional Groups 



Characteristics 



Taxa 



Filamentous-Group 



Sheet-Group 
Coarsely-Branched-Group 



Thick-Leathery-Group 



Jointed-Calcareous-Group 



Crustose-Group 



Thin, uniseriate, 

multiseriate 

or lightly corticated 



Uncorticated, foliose 
Corticated 



Differentiated, heavily 
corticated, thick walled 

Calcified genicula, 
uncalcified intergenicula 



calcified or uncalcified 
parallel cell rows, 
encrusting 



Gelatinous diatom, 
Caulerpa verticillata , 
Dasya rigidula , 
Spyridia f ilamentosa , 
Wrangelia sp. 

Dictyota dichotoma 

Caulerpa mexicana , 
Dictyosphaeria cavernosa , 
Hypnea sp. , 
Penicillus capitatus , 
Penicillus pyriformis . 
Valonia ventricosa , 

Thalassia testudinum 



Amphiroa fragilissima , 

Halimeda monile , 

Halimeda opuntia f . triloba 

Neogoniolithon strictum 



16 



Table 5. Comparative upper production rates of macrophyte 

communities in tropical marine shallow water ecosystems. 



Community type 



Productivity 
g carbon/m z d 



Location 



Study 



Mangrove banks 

channel 17.2 

bay 13.4 

Seagrass Meadows 

Cymodocea nodosa K.D.E. Ktinig 18.7 
dominant 

Syringodium isoetifolis 5.8 

Archers and Graeb dominant 

Thalassia testudinum dominant 12.5 

Thalassia testudinum dominant 16.0 

Thalassia testudinum dominant 9.0 

Carbonate Reefs 

Shallow fore and back reefs, 5-7 
algal turf dominated 

Fringing reef, 3.8 

Neogoniolithon f rutescens dominated 

Intertidal, 0.65-2.15 

blue green algae dominated 



Macroalgal dominated 



1.5 -3.0 



Belize 
Belize 



This study 
This study 



Photosynthetic corals and 1.6-7.2 
algal turfs 



Mediterranean Gessner & 

Hammer, 1960 

Laccadives Qasim & 

Bhattathiri, 1971 

Cuba Buesa, 1972 

Florida Odum, 1963 

Texas Odum & 

Hoskin, 1958 



St. Croix Brawley & 

Virgin Is. Adey, 1977 

French Sournia, 1976 
Polynesia 

Enewetak Bakus , 1976 



Canary Is. Johnston, 1965 
Enewetak Smith, 1973 



< 

Q. 
CE 

LJ 

_l 
3 
< 



<< 

SI? 

XO 



t-<x 

Q<Zi 

UJ<o 

=i<0 
<XLU 
IhZ 



<_l 
x< 



f~ 



m m o 
to 0) o 
ci d d 



in mm mmmm mmmmmm mmm'nmmm 

IDWW'JNOO- I0-0(DO)W in 00 s O * IO ID 

O — • — cvi cvi — ' oJ ro iri co ro" ro s}-' ^ >t' iri w m' (D id 





00 




o 


1— 




LJ 




O 




U_ 


ID 


Ll_ 


O 


Ld 




O 




O 




>- 




1- 


sT 


q: 


O 


< 





(f) 



I 




a 
c 

•r-1 


O 

o 



O 1 - 



3 6 




a" co 




vi 




CD 60 




•u o 




■H u 




CO -o 









,>. CD 




cO t3 




,fi 




cd 




CD 

43 CO 

■u co 


4J 

43 
00 


u -a 




O 01 


>4-l (0 


(11 


(0 

cd aj 

4-> u 


43 

4-1 


CO ft 


a 

o 


CD 


CD r-s 


a 


> co 


> 

•■-I 

00 


O U 


O 0) 
4-1 


CD CD 




00 
CO 


ID 
U 

cd 


A .H 

CD 


CU 


O 4J 


3 
O 

00 


CD CD 
ft CO 


A co 
O M 


n 

CD 

4-1 

co 


CD 00 


CO pj 

CO o 


o 


43 .-i 
cd 

CO 
•i-l CD 

co y 


43 

o 

CO 
CD 


>« C 

1-4 {0 
CO 4J 


C 


f3 co 




CO .H 




T3 


ti 


K 


to 


CD >, 


•7-1 


J-i x> 


a 


CO 


cfl 


s -a 


00 


t— 1 CD 


u 


O i-l 


o 


i-l 




CD 


r-1 


: •*> 


CD 


«N ta 


> 


i— i 


O 


CD w 


a 


H 




P 




00 




•I-l 




fa 





<g 

El 
> — i 
a. < 

<S) x 



< 

a 

Ld 



< 

I 



< 

_i 
< 

x 
i- 



/ V V V N 

OmifiifimifiWinifiifiininininm n m ^m lOiommmmmio mm m m 

OiototnNiDOio^s-iosiDtvj en — oo in roioqioq^sw oj cr> en to 

OOOO— — ro - pJcocJrOirird^' r<"> iri sf' sf' iDuiiDuiicifiifiK NIC t£) (d 

J 



00 

b 




Ld 
O 

Li_ 

Ll 
Lid 

o 
o 



>- 

t * 

q: o 
< 



u~> 



CvJ 

d 



U 



a 


C 




■u 


to 




id 


c 




!-• 


•H 




T3 


fl 




tO 


o 




3 


a 




cr 






a) 


. 




4-1 


B 




•H 


01 




in 


u 
M 




r-l 


o 




<u 


M 




a 


TO 




c 


a 




tO 


01 




43 


■O 




o 


to 




01 

x: 
4-1 


U3 

tO 


4J 

X! 
00 


M 


T3 


•rl 

M 


a 


OJ 


14-1 

tO 
4-1 


tn 
en 
0) 
l-i 


OJ 

A 

4-1 


to 


P. 




-a 


X 

0) 


u 

o 


U 






0J 


^-^ 


u 


> 


tn 


<u 





>-i 


> 


o 


0) 


•H 




4-1 


OU 


<1J 


OJ 




60 


B 


01 


to 




M 


4-1 


c 


tO 


c 


■ w 




11 




u. 


o 


4-1 


3 


M 


tJ 


tj 


0j 


H 


l-l 


o. 


tn 


Hi 


c 


a 


u 


o 


t-l 


OJ 




4-J 


4-1 


X) 




Ul 


01 


M 


3 


tn 


r. 




to 


o 


u 


Xi 


i— i 

(0 


X! 


■ 




U 


■•-> 


a) 


to 


V) 


u 


0/ 


P^ 


c 


G 


r-l 


to 


tO 


4-1 


•H 


a 


tn 




to 


■rH 






■n 


b 


h 




tn 


dj 


>^ 


•H 


4J 


X> 


C 


a 




to 


3 


TO 


OO 


r-l 


01 


M 


U 


r-l 

r- 1 







01 


4-1 


• 


XI 


OJ 


ro 


tO 


> 




r-l 





OJ 


S-/ 


o 


H 






3 






00 








/~s 


O 


en 


.-1 




a. 


m3 






m 


CM 


<u 




a- 


CD 
00 


a- 

3 


■.-( 




fa 


Qi 




4-1 


ti 


• H 


O 


en 


u 




U-t 


>. 




CO 


-o 


XI 


CD 




a 


(11 


•t-i 


X. 


y 


4J 


M 




cu 


■U 


4-1 


m 


CD 




T3 


a) 


"^ 


e 




•H 


U3 


i-H 


U, 


<i> 


3 


m 


O 


n 


l-l 


.a 


00 


CO 



Li_ 


a 




cd 




u 




•i-t 


UJ 


a 


O 


o 


~z. 




ro < 


>4-l 

o 


1- 




CO 


CO 


Q 


o 



<u ^ 



8P 



"3 « 

a. 
a> 

u «-i £ 
o o * 

t-1 U ° 

u *-' 

§ ii » 
S "O 

50 ~ ■ 

H O 

60 *J 
O - 



QJ 

a 
5 5 1 

4J O 

. co 
Si" i-l T3 

ai a 

O) 1-1 CO 

oo 
•l-l 
fa 




AiiAiionaodd 









co 
a) 

4J 

P. 

o 



s3 



c 
co 
c 

B 


CO 

o 

■H 


o 


a 


•o 


> 


<*-t 


>. 




j=i 


/ *- s 




a 

3 

u 


9 

> 

O0 


4J 
CO 

ja 
3 
co 


in 

•H 

B 




•H 


<s 


i-l 


B 


u 




a 


CO 


01 


01 


T3 


■H 


•H 


4J 


U-l 


■H 


c 


> 


o 


• H 


u 


4-1 




o 


M 


3 


Ln 


*o 


cr* 


o 


+ 1 


u 




a. 








>,<=> 


u 


ao 


CO 


Oi 


s 


— i 


•rl 




u 

a. 


-—I 

■r-t 




1-1 




a. 


G 


< 


0> 


<t 


CO 


CN 


o. 

a. 


C 


cd 


'j 


4J 

0) 


co 
>> 


z 


CO 
O 




a 


1/1 


• H 




3 


0) 


H 


M 




3 




00 





ATOLL RESEARCH BULLETIN 
No- 290 



SOME OBSERVATIONS ON NESILLAS ALDABRANUS , THE 
ENDANGERED BRUSH WARBLER OF ALDABRA ATOLL, 
WITH HYPOTHESES ON ITS DISTRIBUTION 



By 
C Hamrler, K. Hambler and J. M. Newing 



Issued By 
THE SMITHSONIAN INSTITUTION 
Washington, D- C, U.S.A. 
May 1985 



CONTENTS 

INTRODUCTION 1 

I : RECENT OBSERVATIONS 2 

1) Observations 2 

2) Discussion of recent observations 3 

II : THE HABITAT OF ft. ALDABRANUS 5 

1) New observations on the vegetation of western lie Malabar 5 

2) Review of features of the classic habitat 6 

a) Extremely dense vegetation 6 

b) Large stands of Pandanus tectorius 7 

c) Abundant Dracaena ref lexa 8 

d) Absence of tortoises and goats 8 

i) Numbers 9 

ii) Distribution 9 

iii) Impact 10 

3) New hypotheses on the habitat of .N. aldabranus 11 

3.1) Other peculiarities of the Gionnet region 11 

e) Relatively high rainfall 11 

f) Relatively species-rich flora 12 

3.2) Synthesis of features of the classic habitat 12 

g) Predicted micro-climate with high relative humidity ... 12 
h) Predicted high invertebrate food supply 13 

3.3) Predicted distribution of N_. aldabranus 13 

III : CONSERVATION OF N. ALDABRANUS 15 

SUMMARY 16 

ACKNOWLEDGEMENTS 17 

REFERENCES 17 



< 

■a 

CO 

•p 



CO 


o 


w 




H 


CO 


t ) 


CD 


Z 


■P 



c 

(U 



■P 

CO 

o 



-p 

o 
c 

CD 



CO 

o 



! 

CO 

a) 



-p 

<D 
Ph 

<D 

c 

< 



>> 
-a 
3 

-p 

CO 



p 
x 
a) 

-p 
to 
a> 

3 



CO 
a! 

0) 

CO 

CD 

p 

O 

c 



■p 

CO 

o 

ft 

■a 

0) 

n 

(1) 
,a 

e 

3 

c 

tp 

o 



3 

C 

co 



o 

CSJ 



c 
o 

•H 
-P 

CD 



Cm 
O 



u 
p< 
p, 
nJ 

co 

o 

CO 



t 

o 

CO 



g 

-p 

CD 

c 

a> 
p, 





w co 

<M 



SOME OBSERVATIONS ON NESILLAS ALDABRANUS , THE 

ENDANGERED BRUSH WARBLER OF ALDABRA ATOLL, 

WITH HYPOTHESES ON ITS DISTRIBUTION 



By 
C. Hambler* K. Hambler and J. M. Newing 

INTRODUCTION 

Nesillas aldabranus Benson and Penny, the Aldabran Brush Warbler, is 
endemic to Aldabra atoll, Republic of Seychelles. It is considered the 
world's rarest, most restricted and most highly threatened bird (Collar 
and Stuart 1985). 

This paper presents relatively recent observations of _N. aldabranus , 
and considers the features of its habitat which may be involved in the 
very restricted distribution of this species on Aldabra. Testable 
hypotheses are presented which could form the basis of future work on this 
species, and which might help in efforts to conserve it. 

The warbler was discovered in 1967 (Benson and Penny 1968), and as 
much of the atoll had already been sampled by mist-netting it was evident 
at that time that the species was not distributed throughout Aldabra. 
Extensive searches by R. P. Prys-Jones, between July 1974 and Feb. 1977, 
revealed five individuals; all were within a 50 m wide, 2 km long strip 
along the northern coast of western lie Malabar near Gionnet (see Fig. 1). 
This strip will be called the Gionnet region, and the vegetation within it 
the classic habitat of _N. aldabranus . No warblers have ever been seen 
away from their classic habitat, but it was hoped that the little-known 
southwestern region of Grande Terre might support a population; however, 
it is now known that the mixed scrub in the SW is dissimilar to that at 
Gionnet (Cowx 1980, D. McC. Newbery pers. comm. and personal observation) 
and limited searches with call playback revealed no warblers (C. Peet 
pers. comm.; personal observation 1983). It is thus considered unlikely 
that a population of N_. aldabranus exists elsewhere than on lie Malabar. 

The population of the warbler was estimated to be at most 25 
individuals in 1977, based on knowledge of behaviour and extrapolation 
into the extent of likely habitat (Prys-Jones 1979). The habitat 
considered suitable was that similar to the mixed scrub near the coast in 
the Gionnet region, and was thought to extend to ca. 1 km west of Anse 
Grand Grabeau. 



* 14 Yew Tree Avenue, Bradford BD8 OAD, England. 



The classic habitat was considered to have four features which, taken 
in conjunction, made it distinct from other areas of the atoll (Prys-Jones 
1979). These were : a) extremely dense, closed-canopy vegetation, with a 

considerable leaf litter/soil layer beneath; 

b) large, dense stands of almost pure Pandanus 
tectorius ; 

c) a high abundance of Dracaena ref lexa ; 

d) a total absence of both tortoises ( Geochelone 
gigantea Schweigger) and goats ( Capra hircus L.) 

In this paper we review these four features, considering recent 
observations of the warbler and using selected data from a study we made 
of the composition and architecture of the mixed scrub of lie Malabar 
(which will be published separately). 

We report here some observations on the warbler and related subjects 
made by the Animal Ecology Research Group, Oxford, Expedition to Aldabra 
(on Aldabra 2 Aug. to 26 Sept. 1981), the Southampton University 
Expedition to Aldabra (11 July to 15 Sept. 1982) and the Cambridge Aldabra 
Rail and Brush Warbler Expedition (19 July to 24 Sept. 1983). 



I. RECENT OBSERVATIONS OF _N. ALDABRANUS 

1) Observations 

Observations of the warbler up to 1977 are given in Benson and Penny 
(1968) and Prys-Jones (1979). Since R.P. Prys-Jones left Aldabra in 1977 
there have been, to our knowledge, four definite records of _N. aldabranus 
and a number of observations of birds thought to be of this species. 
These records are listed in Table 1. It has been pointed out by R.P. 
Prys-Jones (pers. coram.) that very occasionally vagrant warblers of other 
species may be found on Aldabra, and that although the long calls of _N. 
aldabranus are distinct from the vocalisations of other indigenous birds 
(O.E. Prys-Jones, pers. comm.), confusion might conceivably arise with 
such vagrants; therefore, only good visual records can be treated as 
certain. 

We describe some of the recent observations in detail below: 
In 1981, whilst walking along uncut transects, C. Hambler and T.C. 
Guilford heard a very distinctive call, subsequently identified by 
comparison with taped calls as almost certainly that of N_. aldabranus . 
The site was in thick mixed scrub about 250 m east of Anse Petit Grabeau, 
and about 100 m inland close to vegetation dominated by Pemphis acidula . 
The call, lasting about two seconds, comprised a series of clicks in rapid 
succession; the frequency of clicking increased in the middle of the call, 
which was thus described non-phonetically as a "rattle, chirr, rattle". 
This may be the alarm call described by Penny as a harsh "chirrr" and by 
Diamond as a short, scolding chatter (in Benson and Penny, 1968). 

On 1 Sept. 1983 at ca. 5 p.m. a chak sequence (Prys-Jones 1979) was 
heard three times by one of us (K.H.) in a thick Pemphis acidula bush, 
beside the northernmost pebble beach on the Malabar side of Passe Gionnet; 
several attempts were made to find this bird again in the following week, 



including use of call playback and squeaking, but with no success. 

On 2 Sept. in the early afternoon we heard two chak sequences 
between posts A28 and A29 on the coastal path at Gionnet. These calls, 
each lasting about 4 seconds, were probably the "machine-gun chatters" 
described by Prys-Jones (1979) as being given rarely, in excitement. The 
calling bird was seen, and it approached us--possibly attracted to 
squeaking. It did not respond to call playback and examined us from a 
distance of about a metre, before moving off and feeding in the scrub. 
The bird bore a single ring, on the right leg, so must have been one of 
the four birds ringed as adults in 1974; the colour ring on the left leg 
has been lost. Details of ringed birds are given by R.P. Prys-Jones 
(1979 — for colour rings, and pers. comm.--for metal rings). 

A warbler with a single ring, as above, was seen (by G.H.) around post 
A29 on 4 Sept., and between post A27 and A28 on 6 Sept. As other birds 
might have lost colour rings, we cannot be sure that the same individual 
was involved in each case. 

In all the September 1983 sightings the bird was observed feeding 
actively; it appeared in good condition and was not in moult. In the 
sightings on 1 and 4 September the bird took many food items too small to 
identify, and was seen several times to hang beneath twigs and stems while 
pecking at the underside of leaves--such hanging behaviour was not 
observed by R.P. Prys-Jones (pers. comm) . 

2) Discussion of recent observations 

The recent observations are interesting mainly because of the extreme 
rarity of N_. aldabranus ; although they add little to what is already known 
of the species, any information is worth examining in case attempts can be 
made to save the bird, or as the last records of a species becoming 
extinct. 

At least one of the birds ringed between 16 April 1974 and 17 Dec. 
1974 survived in 1983, indicating that an age of at least 9 years can be 
reached. This is not exceptional for a small tropical bird (Prys-Jones 
and Diamond 1984). A mean life expectancy of 8.9 years is given for the 
Seychelles brush warbler ( Acrocephalus sechellensis ) by Diamond (1980), 
and individuals of that species have been observed engaged in breeding 
behaviour at 10 years old (V. Wood, per. comm.) with some surviving to at 
least 11 years (H. Owen, pers. comm.). 

The bird(s) we saw on 2, 4, and 6 Sept. 1983 showed an initial 
interest in us, and then wandered off feeding; in the latter two 
sightings, squeaking and call playback were not used to attract the bird, 
and indeed in the first sighting of a bird on 2 Sept., we were not even 
seeking the bird when it first appeared near us. We had the distinct 
impression that the bird spontaneously approached people as they moved 
noisily through the vegetation on the path. It is therefore interesting 
that curiosity and tameness have been noted in the probable ancestor of N^. 
aldabranus ; the polytypic N^. typica of Madagascar and the Comoro Islands 
(Rand 1936; Benson 1960). It is possible that when approaching in 
"curiosity" the bird is seeking disturbed invertebrates—perhaps also why 



Nesillas is a common member of parties of mixed species on Grande Comoro 
(Benson 1960). As will be discussed later (Section II, 3) invertebrate 
abundance or activity may be crucial to the distribution of N. aldabranus . 

All of the records of N. aldabranus fall within the Gionnet region, 
with the notable exception of that from near Anse Petit Grabeau in 1981, 
which is some 9 km east of the nearest previous record. As this record 
was of voice only, it is subject to the reservation mentioned above. The 
isolated observation of _N. aldabranus here would present many interesting 
possibilities, but information on the species is so limited that 
interpretation is virtually pure speculation. It suggests that the 
warbler can at least survive in habitat not previously considered 
suitable. However, virtually nothing is known of dispersion in this 
species, other than that it may account for periods when individuals 
"vanished" from the study area at times between 1974 and 1977 (Prys-Jones 
1979). 

It is unlikely that a "sizeable" population (i.e. one as large as 
that of the Gionnet region in the 1974-1977 period) could have escaped 
detection in the Anse Petit Grabeau to Anse Malabar area, as this area has 
been visited many times on various research and management projects. We 
visited the area of the 1981 record in 1983, and used call playback and 
squeaking without result. If a population did exist there it would be 
less likely to be discovered if it were more than about 50 m inland, away 
from the coastal path, in more dense and less visited vegetation. 

Our exploration of the Anse Petit Grabeau area in 1981 and 1983 
revealed a mosaic of mixed scrub and patches of Pemphis , with the mixed 
scrub being closer in composition to "Malabar Mixed Scrub" than to 
"Gionnet and Polymnie Mixed Scrubs" (vegetation classification after 
Gibson and Phillipson 1983). Unless other observations of the warbler are 
made in the Anse Petit Grabeau region, or east of Opark, we believe there 
are insufficient grounds to include the vegetation in this region as 
suitable habitat for a population of N_. aldabranus . 

The nest we discovered west of Opark presents the possibility of the 
Gionnet population of N. aldabranus extending to Opark inlet, which would 
be consistent with the observations on vegetation discussed in Section II 
1 below. 

There are not enough data to compare the size of the population of 
warblers near Gionnet in the early 1980's with that in 1974-1977. The 
species is difficult to observe, and responds less readily to call 
playback in the dry season when our visits were made (R.P. Prys-Jones, 
pers. coram.). The recent infrequency of observations does not give an 
indication of a change in abundance, which could have occurred without 
detection. We discuss in the next section changes in the vegetation on 
lie Malabar since the 1970's, which lead us to suggest that the maximum 
likely population is about half that previously estimated, without 
assuming a change in the real density of the species. 



II. THE HABITAT OF N. ALDABRANUS 

1) New observations on the vegetation of western lie Malabar 

The critical factor in assessing the likely distribution limits of N. 
aldabranus on lie Malabar is the distance to which habitat that is 
considered suitable extends east along the northern coast of the island. 
Various differences in the species composition of the mixed scrub 
vegetation of eastern and western lie Malabar have already been described 
(Newbery and Hill 1981; Gibson and Phillipson 1983). A change occurs 
from the Gionnet type of mixed scrub to the more extensive Malabar type, 
but the details of the transitional zones were not known. Certain species 
of plants are exceptionally abundant in the NW of Aldabra ("northwestern 
species": Gibson and Phillipson 1983), and the frequencies of such species 
decline in the eastern region of lie Malabar; we investigated this decline 
in 1983, using uncut transects marked on Fig. 1, to try and discover how 
mixed scrub suitable for the warbler graded into apparently unsuitable 
types. 

Preliminary analysis of our results on the densities of plant species 
in the mixed scrub reveals that for a number of species there is a sharp 
decline in abundance in a particular section of the island: the Opark 
area. At this site there is an interruption of the coastal cliff, with a 
bay about 100 m wide which extends about 200 m inland before narrowing 
into a mangrove creek. As far as we could tell, this separates the mixed 
scrub belts along the northern coast on either side. 

There are many differences in the density, architecture, and species 
composition of mixed scrub near Gionnet and Anse Petit Grabeau (Hambler et 
al. in prep.), but our most significant discovery is that some such 
differences are evident on opposite sides of Opark, despite the relatively 
short length of coastline involved. We illustrate this vegetation change 
in Table 2, in which selected species only are included; there are other 
plant species which do not follow the same trend, but we consider the 
species listed in the table are particularly useful in the context of a 
habitat change which might limit the warbler, and certain of these species 
are discussed in more detail in the following sections. The differences 
in vegetation across Opark, but particularly the dramatic decline in 
Dracaena reflexa and Tarenna supra-axillaris gave the very distinct 
impression that Gionnet- type mixed scrub ends at Opark. 

It is probably not possible to determine how long there has been a 
difference in the vegetation on the opposite sides of Opark; the 
possibility that it is of recent origin cannot be excluded. We know of no 
studies of the vegetation in this area prior to our own in 1983. It is 
possible that the substrate is different west of Opark (which is an 
exceptional feature on the coastline), but a map of the geology of Aldabra 
(Braithwaite et al. 1973) does not suggest this; moreover, the 
topography of the Gionnet region is not thought peculiar (Thorpe and 
Stoddart 1969). We suggest that the vegetation differences across Opark 
may be partly the result of the activity of large herbivores, as discussed 
in Section II 2 d. 



We now review individually features of the classic habitat, and its 
extent. 

2) Review of features of the classic habitat 

Although the combination of the four features listed in the 
Introduction is probably unique to Gionnet, it remains to be explained why 
this combination of features might be important to N.- aldabranus whilst 
other unique habitats on Aldabra are apparently unsuitable. 

a) Extremely dense, closed-canopy vegetation, with much litter and 
soil. "Dense" is a somewhat relative and ambiguous term in the context of 
vegetation, and is often used as a general description of the combined 
influences of rooted-plant density and architectural features such as 
abundance of twigs, leaves, etc. 

The mixed scrub near Gionnet has been described as "dense tall 
scrub about 15 feet high, in places almost forest" (Fosberg in: Benson and 
Penny 1968), and as "extremely thick mixed scrub."... with exceptional 
"thickness of twigs and canopy right down to ground level" (Gibson and 
Phillipson 1983). 

In 1983 we sampled mixed scrub on lie Malabar, examining the 
densities of rooted woody plants, the height of the vegetation and the 
number of twigs of different sizes at various heights. The results of 
this study will be published elsewhere, but the general conclusion was 
that the scrub at Gionnet was not unique in its general density, but it is 
probably more consistently dense throughout the Gionnet region, with fewer 
of the more open areas found in mixed scrub elsewhere on lie Malabar. 
There was a more obvious difference in the architecture of the mixed scrub 
near Gionnet than in the density of rooted plants: the abundance of U. 
ref lexa and Pandanus spp. produced stands of vegetation which were locally 
extremely dense, as these plants have a rather different growth form to 
the other shrubs and trees (see illustrations in Fosberg and Renvoize 
1980). 

A rough relative measure of the density of the scrub is the ease with 
which it can be penetrated by people in a straight line without cutting. 
" Pemphis scrub" (Gibson and Phillipson 1983) is virtually impenetrable, 
and the bulk of the vegetation of lie Malabar is Pemphis , which grades 
into mixed scrub along the northern part of the island. Using this 
measure, density increases inland away from the northern coast; we were 
able to penetrate an average of 225 m due south from the northern coast in 
the Gionnet Study Area, but 350 m in the Grabeau Study Area. In the 
Grabeau area and to the E the denser scrub is generally further inland, _N. 
aldabranus does not appear to use the densest scrub ( Pemphis ). 

The mixed scrub of lie Malabar is heterogenous in many respects, and 
it is possible to find local areas with physical densities similar to that 
near Gionnet further east; the vegetation of lie Polymnie is also very 
dense locally. 

We sampled the height of the vegetation twice every 5 m of our 
transects for the first 100 m from the northern coast. The mean heights 



and standard deviations for the Gionnet study area were: to 50 m : mean 
3.38 m, s = 1.04; 50 to 100 ra : mean 2.77, s = 1.02. For the Grabeau 
area: to 50 ra : mean 1.90 m, s = 1.67; 50 to 100 m : mean 2.51 m, s = 
1.06. This shows a highly significant (p< 0.001) decrease in the height of 
the mixed scrub over the first 100 m from the coast near Gionnet, and a 
significant (p< 0.01) increase in the same distance in the Grabeau area; 
for the full 100 m, Gionnet scrub is significantly taller than that near 
Grabeau (p< 0.001), but this is due to the difference in the most 
northerly section. These data suggest a habitat change inland, and to the 
east. 

N. aldabranus may inhabit the most consistently dense and tall mixed 
scrub on lie Malabar; lie Polymnie appears rather similar, but requires 
study. Although this appears the least unusual feature of the classic 
habitat, it is worth noting that N_. typica on Madagascar prefers thick 
vegetation (Milne-Edwards and Grandidier 1879; Rand 1936) although it is 
not confined to it (Benson 1960). 

The depth of soil and leaf-litter at Gionnet is not unique on lie 
Malabar (pers.obs.) but an unusual shallow organic type is present 
(Trudgill 1979), which may be important. 

b) Large, dense stands of almost pure Pandanus tectorius . 
Pandanus stands were considered to be clearly associated with N_. 
aldabranus from general observation of the birds (R.P. Prys-Jones, pers. 
comra. ) , and the plant is used as nest material and as a site for nesting 
(Benson and Penny 1968). However, the high use the warbler makes of 
Pandanus spp. at Gionnet may be purely facultative, and we believe there 
is insufficient evidence that the distribution of N_. aldabranus is in any 
way related to the distribution of this plant. Prys-Jones (1979) presents 
some data suggested to support such an association, but we believe that 
the single quantitative measure that he uses (linear abundance) is not 
sensitive enough to give an accurate comparison of the amount of Pandanus 
at various sites. We found it impossible to quantify Pandanus abundance 
with any single measure, as the growth form of this large plant produces 
stands which differ considerably in height, shape, orientation and 
"solidness", reflecting the number of leaves and stems and the arrangement 
of s terns . 

Pandanus spp. would be available to N_. aldabranus in the mid and 
eastern regions of He Malabar (and throughout much of Aldabra), but it is 
clear from journeys by boat along the northern coast of He Malabar that 
there is considerably more Pandanus in the western regions; the dominance 
by the species of the Gionnet vegetation has been noted by Gibson and 
Phillipson (1983) as exceptional. Some P. aldabrense was found near 
Gionnet in 1983 (F. Friedman, pers. comm.), but it is probably rare on He 
Malabar (C W.D. Gibson, pers. comra.). 

A possible direct relationship between Pandanus and the warbler 
might be that the plant providas physical protection against predation of 
the nests by rats ( Rattus rattus ) (Prys-Jones 1979). We suggest a 
possible indirect importance of Pandanus in the habitat of _N. aldabranus : 
the structure and height of these plants might provide some form of 
shelter on both a large scale in the Gionnet region, and on a 



micro-environmental scale within the stands. This will be discussed 
further in Section II 3. 

c) A high abundance of Dracaena ref lexa 

The particularly high abundance of D. ref lexa was considered by 
Fosberg to be the only ecological peculiarity of the vegetation in which 
the warbler was first found (Benson and Penny 1968). The dominance of the 
Gionnet vegetation by this species (and by P_. tectorius ) is one of the 
characters Gibson and Phillipson (1983) note as exceptional compared with 
other Aldabra vegetation. D. McC. Newbery (pers. comm.) also considers 
the abundance of _D. ref lexa to be a particularly notable peculiarity of 
the classic habitat. 

D. ref lexa is generally uncommon on Aldabra; its distribution follows 
a "Northwest and Groves" restriction pattern (Gibson and Phillipson 1983). 
It is most abundant in Gionnet type mixed scrub, in which Gibson and 
Phillipson found it to be about ten times as frequent as in the vegetation 
in which it was next most abundant (Gionnet and Polymnie mixed Pemphis ) ; 
lie Polymnie mixed scrub is very similar to that at Gionnet, with the 
important exception that _D. ref lexa is far less abundant. lie Picard has 
very local and limited stands of _D. ref lexa , and the Groves area of 
woodland on Grande Terre has a few plants of this species. 

In our Gionnet study area we found that J), ref lexa is common from the 
northern coast to at least 200 m inland, and is present more than 200 m 
inland here and W of Opark. It is noteworthy that the "Z" path, a 
transect which crosses lie Malabar due south from the Gionnet region, is 
not representative in this respect, since it suggests the belt of high 
abundance of D. ref lexa extends only 50 m inland from the coast 
(Prys-Jones, 1979: Appendix 2). This is an anomaly produced by the Anse 
Porche beach, behind which the path starts; this beach and bay occupy an 
area which would have been mixed scrub if the coastline had been straight; 
atypical plant species and human disturbance here are further 
complications. We found ]D. ref lexa with high abundance, in thick stands, 
immediately W of Opark, and it is probable that it is present in similar 
quantity throughout the mixed scrub/mixed Pemphis between Opark and Passe 
Gionnet. 

The northwestern distribution of J), ref lexa (with certain other 
plants) may be explained by their sheltered position relative to the dry, 
salt-laden SW trade winds (Gibson and Phillipson 1983). The leaf surface 
of JJ. ref lexa is not protected against high water loss (D. McC. Newbery, 
pers. comm.) and the species appears better adapted to the wetter parts of 
Aldabra. This will be discussed further in Section II 3. 

N_. aldabranus is known to spend considerable time foraging in stands 
of J), ref lexa , although there is not sufficient evidence to demonstrate 
selective use of this plant (Prys-Jones 1979). We predict that detailed 
study would confirm such an association, since we predict certain 
invertebrates will favour the moist micro-environment this plant is likely 
to provide; this will also be expanded in Section II 3. 

d) The absence of tortoises and goats 

The Gionnet region, and lie Polymnie, are unusual in that local 



populations of the endemic giant tortoise and feral goats have recently 
been very low or absent (Bourn & Coe 1978; pers.obs.). These large 
herbivores may indirectly influence the distribution of N,. aldabranus and 
so we present here evidence of population changes in these species which 
we suggest are of significance. 

i. Numbers 
Tortoises were reduced to very low numbers by exploitation before 
this century, and populations easily accessible from the settlement at lie 
Picard were virtually exterminated; their numbers recovered rapidly this 
century when exploitation declined (Stoddart and Peake 1979). The 
population of tortoises on lie Malabar was estimated to be about 2000 in 
the early 1970's (Bourn and Coe 1978), whilst there is not thought to be a 
population on lie Polymnie (C.W.D. Gibson pers. comm.). 

Goats were introduced to Aldabra before 1890, and their total 
population has suddenly increased alarmingly; in 1977 the total number was 
estimated to be 500 to 600 (Gould and Swingland 1980), but in 1982 it was 
estimated at 2560 + 560 (Newing et al. 1984). The population on He 
Malabar was estimated to be 200 to 250 in 1976 (Gould 1979), and that of 
He Malabar east of Anse Grande Grabeau was estimated to be 289 + 30 in 
1982 (Newing et al. 1984). In 1983 we found fewer goats at the eastern 
end of He Malabar than in 1982, but more in the Anse Malabar/Anse Petit 
Grabeau area than in 1981. Given the distribution records below, these 
figures show a population increase and spread on He Malabar since 1976. 

ii. Distribution 
Tortoises were probably present throughout the length of He Malabar 
before their exploitation, and were probably particularly reduced in the 
more accessible western areas of the island. Fieldwork in 1972-74 
suggested that tortoises were distributed in mixed scrub from the eastern 
tip of He Malabar to about 5 km E of Opark (Bourn and Coe 1978). Studies 
of goats betweem 1976 and 1977 suggested these also occurred over this 
range (Gould and Swingland 1980), i.e. up to Anse Petit Grabeau. No signs 
of goats or tortoises were found in brief explorations up to about 2 km W 
of Anse Petit Grabeau, nor near Anse Cedres Opark in 1975-76 (I.R. 
Swingland, pers. comm.). Brief visits to Opark between 1974 and 1977 
revealed no signs of these animals (R.P. Prys-Jones 1979). Although 
Seychellois guides and workers think both goats and tortoises were present 
around Opark in the middle of this century (H. Charles, E. Constance and 
R. La Fontaine, pers. comm.), the numbers of large herbivores W of Anse 
Petit Grabeau must have been very low up to 1977. 

In 1981 several goat and tortoise faeces were found about 1 km W of 
Anse Petit Grabeau (C.H., pers. obs.), and in 1983 we discovered goats and 
tortoises were present in numbers far too high to have been overlooked, up 
to the eastern bank of Opark. Tortoises were found throughout the Opark 
study area — i.e. up to 500 m W of Opark, but we could find no signs of 
goats W of Opark. The present western distribution limits of tortoises 
and goats are not clear, although we found no signs of them up to 2 km E 
of Gionnet. It is likely that there are more tortoises than goats W of 
Opark: the number of goats would have to be very low since no faecal 
pellets were found near the coast. There is an unconfirmed report that 
four tortoises were observed near Gionnet Camp in early 1983 (P. Bijoux, 



10 

pers. comm.). Our observations suggest very few Large herbivores were 
present W of Opark in 1983. 

It is clear that there has been a considerable increase in the 
numbers of both goats and tortoises in the western region of lie Malabar, 
particularly E of Opark. This increase is likely to be a result of 
migration from more eastern regions, with continuous populations from 
Opark to Passe Houareau. In 1983 we found that three of the tortoises 
examined near Anse Cedres Opark had disk marks indicating they had moved 
from the Anse Malabar region (some 7 km away) since 1974 (Bourn and Coe 
1978), and tortoises and goats were evident throughout the Grabeau study 
area in 1983. One tortoise nest was found near Anse Cedres Opark in 1983, 
but the smallest tortoises we found were about 10 to 15 years old--also 
suggesting movement was more important than reproduction in the population 
increase E of Opark. 

Opark inlet is dry at low spring tides, and it is easy to cross on 
foot, small bays on each side providing points where access to and from 
the land surface is possible; the western bank is the more difficult to 
ascend and penetrate. The inlet may provide a relatively large obstacle 
to the movement of large herbivores on lie Malabar, but there seems to be 
no absolute barrier, and it is possible that they could move round the 
inlet to the south. Although these herbivores may not favour the thick 
vegetation W of Opark, we believe it is only a matter of a few years 
before they cross the inlet in significant numbers. 

iii. Impact 
Tortoises and goats-- the only large herbivores on Aldabra--may be 
responsible for some of the vegetation patterns on the atoll; however, the 
origins of most patterns are complex (Newbery and Hill 1981; Gibson and 
Phillipson 1983). It is useful to consider their potential impact on the 
habitat of N_. aldabranus , by examining their distribution and behaviour in 
relation to the limits to the likely habitat of the warbler. 

It is possible that the relative abundance of some plant species in 
the NW of Aldabra is related to the degree of exploitation of tortoises in 
these areas. However, exclosure experiments suggest that it is unlikely 
that release from grazing pressure alone would result in vegetation like 
that of lie Polymnie (Gibson et al. 1983), nor like the similar Gionnet 
region. 

In view of the dramatic vegetation differences across Opark, it is 
interesting to consider the known diets of the two herbivores. JD. ref lexa 
is a particularly important species to examine, as are other common 
species of the Gionnet mixed scrub which show a marked decline across the 
inlet. 

Tortoises will readily eat _D. ref lexa (Grubb 1971) and are likely to 
damage it physically. However, they have not eliminated it on lie Picard, 
or in the Groves, and are co-existing with it W of Opark (although they 
were found to be favouring stands of the species, judging by their 
faeces). Euphorbia pyrifolia and Tarenna spp. are considered unpalatable 
species for tortoises (Grubb 1971). It is thus unlikely that tortoises 
alone are responsible for the change across Opark. 



11 



Goats are capable of exploiting most vegetation types on Aldabra, and 
both graze and browse (pers.obs.). Although previously less dominant 
(Gould and Swingland 1980), goats are now having a striking impact on the 
vegetation of Aldabra, and on Grande Terre 36% of the scrub cover of some 
regions has been lost since 1978, particularly those species most 
palatable to goats (Newing et al. 1984). On eastern and middle lie 
Malabar goats are now beginning to degrade vegetation (pers. obs. 1983). 
Goats are reported to take J), ref lexa "more than occasionally" in studies 
of their food preferences on lie Picard (Stevenson 1972); JE. pyrifolia is 
a "preferred" food; Tarenna supra-axillaris and T_. verdcourtiana , however, 
are not eaten at all frequently (but might be indirectly affected by scrub 
degradation) . 

It is probable that goats are. involved in the vegetation change 
across Opark, but they cannot be considered entirely responsible, as some 
less palatable species show a decline across the creek, whilst some 
preferred food plants (such as Polysphaeria multiflora ) did not decline 
greatly. It is likely that a "natural" decline in "northwestern" plant 
species on lie Malabar has been exaggerated by goats. 

We conclude this review of the likely significance of tortoises and 
goats to the distribution of N. aldabranus by suggesting that large 
herbivores may adversely modify the vegetation by selective removal of 
some plant species and by opening up the scrub cover; this may have 
restricted the range of N. aldabranus in the past. We suggest that the 
west/east decline in certain plant species across Opark and the decline in 
large herbivore numbers (particularly goats) in the reverse direction is 
not coincidental. The absence of large herbivores is probably a key 
feature of the classic habitat, but not the most important. 

3) New hypotheses on the habitat of If. aldabranus 

3.1) Other peculiarities of the Gionnet region 
Although the Gionnet to Opark mixed scrub region appeared 
subjectively distinctive to us, it is hard to single out the 
characteristics which gave this impression. The four features reviewed 
above are each fairly obvious at Gionnet, but they do not provide 
sufficient information to explain the limits of the population of the 
warbler. Other areas on Aldabra possess combinations of these features, 
apparently without supporting warblers, and although it is possible the 
birds do not occupy all suitable habitats on Aldabra, it is instructive to 
consider other characteristics of the Gionnet region which add to its 
distinctiveness. These features will be labelled: e) and f). 

e) Relatively high rainfall 
A feature of the northwest of Aldabra that has recently been 
confirmed is that it receives a relatively high rainfall; several years of 
records from rain-gauges at up to 14 sites around Aldabra have been 
compiled by Stoddart (1983), and these are sufficient to detect some 
trans-atoll variations in rainfall in the study period (1973-1981) 
although some sites lack records in some years. 



12 

The NW of Aldabra generally receives a higher mean annual rainfall 
than the other regions of the atoll. In the NW, Anse Var (on lie Picard) 
received the overall highest mean annual rainfall (1567 mm); Gionnet 
received the second highest (1448 mm) and the highest mean annual rainfall 
in three out of the seven years for which records are available (in 1976, 
1979 and 1981); Anse Var received the highest mean annual rainfall in two 
of the four years for which records are available (in 1978 and 1980). The 
rainfall in the NW and at Gionnet is also relatively high in the driest 
six months of the year (June to November), and in this period Gionnet 
receives a consistently high rainfall-- no months with a mean rainfall of 
zero. The more easterly areas of lie Malabar received a lower overall 
mean rainfall in the driest six months than did Gionnet and the NW 
(Stoddart 1983, figure 17). 

Gionnet is probably not exceptional in its local rainfall, relative 
to the rest of the NW, but we consider the high rainfall it receives 
throughout the year to be an important link in the range of features which 
lead to the selective use of Gionnet by the warbler. Rainfall may be 
directly or indirectly important to _N. aldabranus , particularly as in the 
dry season there is very little standing fresh water other than puddles 
after rain. Aldabra is relatively arid by comparison with other habitats 
used by Nesillas species, with the exception of _N. typica lantzii of the 
subdesert in the SW of Madagascar. It is possible that Aldabra is a 
marginal habitat for Nesillas , in terms of water balance as in the dry 
season such a small island might provide fewer opportunities to acquire 
water (directly or in food) than do parts of the SW of Madagascar. 

f) Relatively high species richness of the flora 

The mixed scrub of Gionnet has a notably rich woody flora (D. McC. 
Newbery, pers. comm. ; Gibson and Phillipson 1983). It is again not 
unique--the floras of lies Picard and Polymnie are richer--but this may 
contribute to the quality of the habitat near Gionnet directly (e.g. 
through structural diversity) or indirectly (e.g. through herbivorous 
invertebrates see Frith 1979). 

3.2) Synthesis of features of the classic habitat 
We now combine the known features of the classic habitat ("a" to "f" ) 
with some predicted features ("g" and "h" below) to produce a general 
theory which may explain the restricted range of N_. aldabranus on Aldabra. 

We believe the most important and exceptional feature of the classic 
habitat is the high abundance of Dracaena ref lexa . The distribution of 
this plant is not fully understood, and further examination of the 
substrate near Gionnet would be helpful. Substrate, available water and 
the activities of large herbivores may limit this plant's range on 
Aldabra; as this species is suspected to be susceptible to high 
transpirational water loss, we predict a likely feature of the Gionnet 
region: 

g) Predicted micro-climate with high relative humidity and 
still air 

We suggest that abundant D. ref lexa will both depend on and produce 
a high relative humidity around the stands. This would be facilitated by 
a high and constant rainfall, a topographically sheltered position on the 



13 

atoll, and local windbreaks such as tall mangrove and Pandanus stands. 
Our subjective observations at Gionnet did suggest such a micro-climate, 
which might be important to the warbler in critical periods such as the 
dry season or when rearing young if standing water is less rapidly lost 
through evaporation. We predict a possible indirect importance of such a 
micro-climate: 

h) Predicted high invertebrate food supply 
Rainfall is probably related to the annual abundance and the types 
of insects on Aldabra (Cogan et al. 1971); local rainfall is almost 
certainly related to the local abundance of insects at various sites and 
times on the atoll (Frith 1979). Rainfall may act through increased 
humidity, and might act directly to increase invertebrate abundance and 
activity through decreased water-stresses, or indirectly e.g. through 
delaying leaf-fall. The classic habitat might provide diverse and 
numerous invertebrates for N^. aldabranus , and this might be further 
increased by a rich flora (Frith 1979) and by the absence of strong winds. 
Spiders, winged ants and small moths are known to be eaten by _N. 
aldabranus , at least occasionally in some numbers (Benson and Penny 1968). 
These invertebrates are not sufficiently known on Aldabra to test for an 
exceptional abundance near Gionnet; soil and litter arthropods do not 
appear exceptionally diverse or abundant near Gionnet (Spaul 1979), nor 
did those invertebrates sampled by Heath light traps (Frith 1979). 
However, neither of these surveys is comprehensive enough to support or 
refute our prediction: substrate is probably a major factor influencing 
ground-living invertebrates; the Heath trap is not a satisfactory sampling 
method for micro-lepidoptera (P.H. Sterling, pers. comm.) and many other 
groupfs; and Frith's trap at Gionnet was in "poor mixed scrub". 

The invertebrates likely to be most significant to the warbler are 
those of the appropriate size range, active by day or sheltering by day in 
accessible places such as on bark or on leaves. We suggest the 
appropriate sampling method to test for feature "h" would be beating, and 
the dry season might be the most interesting time to compare the classic 
habitat with other areas. In passing, it is worth noting that there is a 
notorious abundance of mosquitos at Gionnet. 

In concluding this synthesis, we suggest that it is significant that 
Nesillas warblers of the east and northwest of Madagascar "recherchent de 
preference les endroits humides" (Milne-Edwards and Grandidier 1879). We 
suggest that features "a" to "h" could explain the limitation of a species 
of bird with very particular habitat requirements to the Gionnet to Opark 
mixed scrub, in addition this area of the atoll is, objectively, one of 
the most likely places to find a member of the genus Nesillas on Aldabra. 

3.3) Predicted distribution of _N. aldabranus 

If we assume that the features of the classic habitat discussed above 
are in combination important to the warbler, we can predict the likely 
distribution limits to the current population. In general, we would 
predict that if N_. aldabranus is present elsewhere on Aldabra, it is more 
likely to be in the mixed scrub in the NW--given the known (and probably 
related) vegetation and rainfall characteristics of that area. 



14 

The western limit to the population of the classic habitat appears to 
be Passe Gionnet (Prys-Jones 1979). This is a channel about 100 m wide, W 
of which the mixed scrub belt of the northern coast changes somewhat in 
character (Gibson and Phillipson 1983), probably reflecting a geological 
change (Braithwaite et al. 1973). Further W is the 600 m wide channel of 
Grande Passe, and west of this the mixed scrub is considerably different 
from that near Gionnet (Gibson and Phillipson 1983, and pers. obs.). 
Although there is considerably less J), ref lexa on lie Polymnie, we find it 
rather suprising that warblers have not been found there, and this island 
seems the most likely site for a population of the warbler outside lie 
Malabar. It is possible that the absence of continuously abundant JJ. 
ref lexa on lie Polymnie breaks the vegetation into patches of suitable 
habitat too small to support a viable population of warblers. It is also 
posTible that Passe Gionnet provides a barrier to voluntary movement of 
the warbler; it is notable that Nesillas is reported to fly poorly, and to 
be incapable of crossing "grands espaces d'une traite" (Milne-Edwards and 
Grandidier 1879). 

The eastern limit of the Gionnet population seems less likely to be 
related to physical interruptions of the habitat; Opark inlet interrupts 
the mixed scrub, but Pemphis scrub behind it might permit movement round 
it. We suspect that a combination of the decline in rainfall to the E, 
and vegetation gradients, would have restricted the warbler to the western 
parts, and that the change across Opark marks the most likely eastern 
limit to the current population. Our limited sampling suggests that J). 
ref lexa may decline slightly in abundance between Gionnet and the western 
bank of Opark (see Table 2), and it is likely that it declined gradually 
to the east of Opark before the encroachment of large herbivores. The 
nest found in 1983 gives hope that the warbler population does indeed 
reach Opark, but the probable record from near Anse Petit Grabeau in 1981 
is too isolated to seriously suggest that the population extends this far 
~east into such different vegetation. 

Between Gionnet and Opark the mixed scrub belt is some 200 m wide; 
the species richness of the scrub, and the structural influence of 
Pandanus , decline inland from the northern coast. We find it surprising 
that N^. aldabranus was not found more than about 50 m from the coast 
(Prys-Jones 1979, and pers. obs.), and it would be interesting both to 
search intensively for N. aldabranus in mixed scrub between Gionnet and 
Opark, and to examine micro-climatic gradients inland from the coast 
within the mixed scrub; there are patches of mixed scrub between 50 and 
200 m from the coast which closely resemble the classic habitat in most 
features (other than Pandanus stands, few of which occur more than 100 m 
from the coast here), and these seem the most likely marginal habitat for 
the warbler. 

The population of _N. aldabranus was estimated to be under 25 
individuals if suitable habitat extended 50 m inland and 9 km along the 
coast from Passe Gionnet (Prus-Jones 1979). We believe that in 1983 
habitat that was closely similar to the classic one extended only to 
Opark, i.e. 4.5 km E from Passe Gionnet, and using the same assumptions 
for extrapolation the maximum likely population was about 13 individuals 
in 1983. 



15 
III. THE CONSERVATION OF N_. ALDABRANUS 

_N. aldabranus is clearly endangered merely through its extremely 
limited distribution. It is not clear how long the population has been so 
small, but it is unlikely that a very small isolated population could 
survive long enough to become an endemic species with signs of island 
adaptations. It is probable that the warbler was either far more numerous 
and widespread on Aldabra in the past (perhaps in wetter periods of its 
history) or that the population is a relict of a recent colonisation by 
this species from an as yet undiscovered (and possibly extinct) population 
on another Indian Ocean island. If the former suggestion is correct we 
may now be witnessing the natural extinction of a species due to large- 
scale climatic and vegetation changes on Aldabra; long-term climatic 
changes are known to have occurred in the region (Stoddart and Walsh 1979; 
Walsh 1984) and standing fresh water has been more extensive on Aldabra in 
the past (Stoddart et al. 1971). In this case there is probably little 
that can be done to protect the species. If the population has always 
been restricted to a small area of Aldabra, or has been restricted at 
least in the last few centuries, then it is possible that Man might have 
indirectly influenced its abundance through the exploitation of tortoises 
and the introduction of goats and rats ( Rattus rattus ). It is possible 
that the warbler was able to colonise Aldabra, or temporarily flourished, 
because reductions in the populations of tortoises on certain islands of 
the atoll allowed vegetation similar to the classic habitat to expand. In 
the latter case, there would be more chance that the warbler could survive 
in the long term, provided reduction in its habitat through the 
encroachment of large herbivores can be prevented. 

There is some recent evidence of vegetation changes in the Gionnet 
region since 1968: M.J. Penny (pers. comm.) reports that in 1984 he found 
that the areas of mixed scrub at both ends of lie Malabar appeared less 
dense, with more dead twigs and branches, than expected. This would be 
worth investigating further. There is also evidence that the rat 
population is still widespread on lie Malabar (we saw them frequently in 
all study areas in 1983) and there is evidence that their impact on lie 
Malabar is increasing: we found rats had stripped bark from branches of 50 
% of the Mystroxylon aethiopicum we examined, and a few Sideroxylon inerme 
bushes had also suffered damage. This stripping of bark was not apparent 
in the late 1970's (D.McC. Newbery, pers. comm.) nor in 1981 (pers. obs.). 
It is likely that rats will evolve to exploit the fauna and flora of 
Aldabra more seriously. Local control of rats might be possible in small 
areas (e.g. if a nest of the warbler were found in use) using new, highly 
specific poisons (I.R. Swingland, pers. comm.). 

It is vital that the remaining habitat of N. aldabranus be monitored 
and protected. It seems likely that goats and tortoises would degrade the 
classic habitat; the control of goats is very urgent, and easier to 
justify than the control of the native tortoises. Although complete 
eradication of the goats would be extremely difficult (considering the 
history of failures in the attempts to eliminate them from other islands) 
it might be possible to achieve this ideal using extremely specific 
biological control agents such as viruses; this method was suggested to 
control mammalian pest species in the Galapagos (Swingland in Mitchell 
1981). In the short-term, culling of goats in the Opark area, and towards 



16 

Passe Houareau, might reduce the rate at which they are spreading towards 
the west and the classic habitat of the warbler. 

We suggest that human disturbance should be minimised between Passe 
Gionnet and Anse Malabar (to include the area of the 1981 record); any 
tourists visiting the Gionnet region should be careful to reduce the 
possibility that they may introduce exotic seeds or pests. 

Positive measures to encourage the remaining population of _N» 
aldabranus seem theoretically possible if our hypotheses are correct. It 
might be fairly easy to increase the availability of standing fresh water, 
using water-dispensing hoppers; this might be particularly worthwhile in 
the drier months. The food supply might be increased locally if barrels 
of water, with netting over the top, were provided before and during the 
breeding season. Such measures might influence other species, including 
potential predators and competitors of the warbler, and so would require 
very careful monitoring. 

SUMMARY 
No more than one individual of Nesillas aldabranus has been seen at a 
time in each observation of the species since 1976, and there is no proof 
that more than one survived in 1983 although some observations suggest 
that this was possible. 

Records of ringed birds since 1977 include at least one seen in 1977, 
and three sightings of a bird, or birds, which had lost the colour ring 
but bore only the metal ring in September 1983. There were three 
sightings of the warbler in late 1983, in which the view was not good 
enough to see if rings were present; there were no observations of the 
species in 1984. 

Sightings of ringed birds show one bird to have lived at least nine 
years by 1983. 

All sightings of the species have been within the Gionnet coastal 
region of Aldabra Atoll. One probable record, which cannot be proven, as 
only the call was heard, was reported outside the classic habitat in the 
vicinity of Anse Petit Grabeau; this suggests it may be possible for the 
warbler to venture into habitat dissimilar to that at Gionnet. 
Exploration of the vegetation east of Gionnet revealed a very sharp 
decline in the abundance of certain plant species across the Opark inlet, 
most notably that in Dracaena ref lexa . The likely extent of habitat 
considered suitable for a population of the warbler is thus re-defined to 
half its previously suggested length, so reducing the predicted maximum 
population of N_. aldabranus to 13 individuals within the mixed scrub 
between Passe Gionnet and Opark. The extremely restricted range of the 
warbler is discussed; known, recently-confirmed and predicted features of 
the habitat of the species are united in a general theory which helps 
explain this range the abundance of Dracaena ref lexa is considered the 
most ecologically atypical feature of the Gionnet region, and it is 
suggested to be no coincidence that the other exceptional feature of this 
region is that it is the only area of Aldabra which supports a warbler of 
the genus Nesillas. 



17 

The following hypotheses are presented which inter-relate the 
features of the habitat of the warbler and its restricted distribution: 

i) Dracaena ref lexa is a particularly important feature of the 
habitat of N^. aldabranus . 

ii) Relatively high rainfall and shelter from the trade winds 
favour J), ref lexa at Gionnet; this plant may be restricted to the west of 
Gionnet by substrate changes, and to the east by large herbivores and 
lower rainfall. 

iii) A combination of high rainfall and a particularly high 

degree of shelter from drying winds (due to location and tall vegetation) 

promote a high relative humidity in the micro-climate of the classic 

habitat, with the greatest humidity in and near stands of D^ ref lexa . 

iv) High rainfall and a predicted high humidity are significant 
to the warbler for part or all of the year; these features may be 
important through the water requirements of the birds, or through the 
activity, abundance or species composition of their invertebrate food 
supply. 

Large herbivores have spread at least 4 km west towards the habitat 
of the warbler since 1977. Goats are particularly likely to have degraded 
possibly suitable habitat, and will encroach into and degrade the habitat 
of the warbler unless urgently controlled. Rats still present a 
particularly great threat. Some positive conservation measures may be 
possible, and are required immediately if IJ. aldabranus is to be saved. 

ACKNOWLEDGEMENTS 

We were very fortunate to receive much help and encouragement from Dr C. 
W. D. Gibson and Dr D. McC. Newbery, both on and off Aldabra. We thank 
Dr R. P. Pry*s-Jones and Dr I. R. Swingland for very useful comments on our 
work. J. A. Stevenson, F. H. Drinkwater, V. Wood and H. Owen provided 
valuable data, and Dr C. J. Harrison and M. P. Walters kindly examined the 
specimen of a nest for us. Particular thanks go to the Seychelles Islands 
Foundation, and we are indebted to L. U. Mole, J. Collie and the 
Seychellois on Aldabra for making our work possible. Many groups and 
individuals supported the three expeditions from which our results derive, 
and we are most grateful for their help and interest. We thank also the 
members of these expedition teams, and J. A. Hambler for vital typing. 

REFERENCES 

Benson, C. W. 1960. Birds of the Comoro Islands. Ibis 103b: 5-106. 

Benson, C. W. and Penny, M. J. 1968. A new species of warbler from the 
Aldabra Atoll. Bull. Br. Orn. Club 88: 102-108. 

Bourn, D. and Coe, M. 1978. The size, structure and distribution of the 
giant tortoise population of Aldabra. Phil. Trans. R. Soc. Lond. B 
282: 139-175. 



18 

Braithwaite, C. J. R. , Taylor, J. D. and Kennedy, W. J. 1973. The 

evolution of an atoll: The depositional and eroslonal history of 
Aldabra. Phil. Trans. R. Soc. Lond. B 266: 307-340. 

Cogan, B. H., Hutson, A. M. and Shaffer, J. C. 1971. Preliminary 

observations on the affinities and composition of the insect fauna of 
Aldabra. Phil. Trans. R. Soc. Lond. B 260: 315-325. 

Collar, N. J. and Stuart, S. N. 1985. Threatened birds of Africa and its 
related islands. The ICBP/IUCN Bird Red Data Book, Part I (3rd 
edition). Cambridge. ICBP/IUCN. 

Cowx, W. D. 1980. Final report vegetation transects SW Grande Terre. 

Report to Royal Society Aldabra Research Committee. (Unpublished.) 

Diamond, A. W. 1980. Seasonality, population structure and breeding 

ecology of the Seychelles brush warbler Acrocephalus sechellensis . 
Proc. 4th Pan-Afr. Orn. Congr. , 253-266. 

Fosberg, F. R. and Renvoize, S. A. 1980. The flora of Aldabra and 

neighbouring islands. Kew Bull, addit. Ser. no 7. London, HMSO. 

Frith, D. W. 1979. A twelve month study of insect abundance and 

competition at various localities on Aldabra Atoll. Phil. Trans. R. 
Soc. Lond. B 286: 119-126. 

Gibson, C. W. D. , Guilford, T. C. , Hambler, C. and Sterling, P. H. 1983. 

Transition matrix models and succession after release from grazing on 
Aldabra Atoll. Vegetatio 52: 151-159. 

Gibson, C. W. D. and Phillipson, J. 1983. The vegetation of Aldabra Atoll: 
preliminary analysis and explanation of the vegetation map. Phil. 
Trans. R. Soc. Lond. B 302: 201-235. 

Gould, M. S. 1979. The behavioural ecology of the feral goats of Aldabra 

Island. Ph. D. dissertation, Duke University, North Carolina, U.S.A. 

Gould, M. S. and Swingland, I. R. 1980. The tortoise and the goat: 
Interaction on Aldabra Island. Biol. Conserv. 17: 267-279. 

Grubb, P. 1971. The growth, ecology and population structures of giant 
tortoises on Aldabra. Phil. Trans. R. Soc. Lond. B 260: 327-372. 

Milne-Edwards, A. and Grandidier, A. 1879. Histoire physique, naturelle 
et politique de Madagascar, Volume XII, Histoire naturelle des 
oiseaux. Paris. 

Mitchell, A. W. 1981. Operation Drake: Voyage of discovery. London, Severn 
House Publishers Ltd. 

Newbery, D. McC. and Hill, M. G. 1981. Numerical description of "mixed 
scrub" vegetation on Aldabra Atoll. Atoll Res. Bull. 246: 1-13. 



19 

Newing, T. J., Daly, K. H. and Hambler, K. 1984. Feral goats on Aldabra. 
Final report, Southampton University Expedition to Aldabra, 1982. 
Unpublished. 

Prys-Jones, R. P. 1979. The ecology and conservation of the Aldabra brush 
warbler Nesillas aldabranus . Phil. Trans. R. Soc. Lond. B 286: 
211-224. 

Prys-Jones, R. P. and Diamond, A. W. 1984. Ecology of the land birds on 

the granitic and coralline islands of the Seychelles, with particular 
reference to Cousin Island and Aldabra Atoll. In: Stoddart, D. R. 
(ed). 1984. Biogeography and ecology of the Seychelles Islands. The 
Hague, Dr W. Junk Publishers. 

Rand, A. L. 1936. The distribution and habits of Madagascar birds. Bull. 
Amer. Mus. Nat. Hist. 72: 143-499. 

Spaul, V. W. 1979. Distribution of soil and litter arthropods on Aldabra 
Atoll. Phil. Trans. R. Soc. Lond. B 286: 109-117. 

Stevenson, J. A. 1972. A final report on feral goats, from Junior Staff 
Scientist to Scientific Co-ordinator. Aldabra Station Records. 
Unpublished. 

Stoddart, D. R. 1983. Spatial and temporal variablility of rainfall on 
Aldabra Atoll. Atoll Res. Bull. 273: 223-246. 

Stoddart, D. R. and Peake, J. F. 1979. Historical records of Indian Ocean 
giant tortoise populations. Phil. Trans. R. Soc. Lond. B 286: 
147-161. 

Stoddart, D. R. and Walsh, R. P. D. 1979. Long term climatic change in 
the western Indian Ocean. Phil. Trans. R. Soc. Lond. B 286: 11-23. 

Stoddart, D. R. , Taylor, J. D., Fosberg, F. R. and Farrow, G. E. 1971. 
The geomorphology of Aldabra Atoll. Phil. Trans. R. Soc. Lond. B 
260: 31-65. 

Thorpe, W. H. and Stoddart, D. R. 1969. The Aldabra warblers. New Sci. 
41: 649. 

Trudgill, S.T. 1979. The soils of Aldabra. Phil. Trans. R. Soc. Lond. 
B 286: 67-77. 

Walsh, R. P. D. 1984. Climate of the Seychelles. In: Stoddart, D. R. 

(ed). 1984. Biogeography and ecology of the Seychelles Islands. The 
Hague, Dr W. Junk Publishers. 



o o 
co o 



E5 
O 



CD 
Ti 
CO 



00 
[>- 
0\ 



u 

ft 



1 00 
1 ON 



-P tJ 
O C 

C -H 
ft 



•H -P 
X- 
[0 

-p -p 
d ft 

x. e 


-■P 

-Tj -P 

u <& 



S3 n3 

ft 0) 
H > 
nj 
o co 



Xi 

nj 
Xi 
o 

ft 



x: 
-p 



-p 




-P 


CD 


x> 


CD 


§ 


3 


a 


3 


S 


a 





o 


o 


ft 


09 


■ H 


o 




o 



•p -p 

CO CD 

nj g 



o 

O 0) 
CM M 

w 

• $ 

O ft 



K _ 


a 


s § 


o 


k co 


S 


8 H 


g 


g o 


g 




03 


n 




SB 


< 


u 


"^ 



CO 

£9 
o 



ft 1 



w « 



M ft 

nj CD 

H H 
o x> 



•H CD 

XI rH 



^8 
nj ft 
CD 

X! — 

xl 

tO -P 



nj 
O « 



U K 
ft g 



CD 



en 

CO 

on 







■3 

Xl 






CO 

K 

ft 




t3 o 

a c 

nj 



CI bO CM 

CD 

CD H fl 

to o 

-p 

_, x: w 

'JJ M nj 
H .H 

a 

ft O 



45 

H 

CD 
O 



CK 

CM 

< 

-P I 

CD CO 



Si 



CN <TN 

00 00 

ON On 



On 
<M 



<8 



• 8 

« in 



CO co u o 

K « fcj co ! 

W W ft ft H 

ft ft CO < ft 



•H -P 

H ft 

CD 

•• ft 

Kg 

nj O 
CD 
X! CD 00 00 



On On 



ss 



$ s 

u u 

CD CD 

O O 



CO 

CM 

< 

I 

ON [>- 

Si Sj 



s 

o 

CO ft 

o 
X! 

-P CO 
■H -P 
S CO 

CD 



ft 

XI O 

to 

•H CO 

-P -P 

•H 10 

H CD 

FP C 

G XI 

■ri -P 

•H 

TJ £ 
CD 
H 



C 
CD -H 
ft -P 
H -H 
Cl3 H 
ft nj 



•h a 

S Tj 



-•h xi a) nj 

-d B h to 

ft - H s! 2 _• 

o to s s a 



-p >> 

XI -P 
•H H rH 

ft -h ri 

to XI o 
>>'m -p 

H tO CO 
ri O CD 
O ft G 



o 



XI Tj 

-PCS 

■ri h Cj 



* 



ft o 
O »n 



en 

n n co 

00 CO ON 
On On ft 



ON 



ffi ffi g « pr, K 

g a . I g a 

a as a a a 



M O O O O 



TJ - CD 

c M xi 

nj <D -P 

ft o 

<D nj G 

CO S (S 



to 

CD CO 
nj 



to 

.. Q) M 

-p ft -g 

•H nj CM 

X - 
®"g o 

S -| tJ 

3 S CD 
• HO' 

ft o „. 

•H rH T3. 



ft XI 

o o 



-P - H 

ftft 

•H H O 

c^g fl 
to -P 
CD ft G 

■8 o,o 



£ to o 



^a^ 



CD XI 

O to 



CD H 
fH cti 



X! 


o 


-P 


Ht 


.. nj VD 


< CD XI 


o 


•• e n 


-p 


O 





o ft 


§ 


o 
in 


o 


to 


•H 


« c 


o 


o < 



CTN 
00 
ON 



K 10 
ft ft 



ft 



oo 

ON 



-p 

S ft 

nj 



■a 

•H 



cn 

00 

On 
iH 



CTN 
CO 
ON 



en 

00 
ON 



o 

CM 



C*N, 

CO 

ON 



CTN, 
00 
ON 



c 
o 


nj 


o 


to 







nj 


XI 


to 
nj 


a 


H 


C 





nj 
o 











to 


to 


a 




C 


•H 




■ H 


$ 





4§ 


H 


rri 


U 





Xl 





O 


ttj 

Xl 


o 


-P 


o 


•p 


to 


H 


CO 


o 


ft 


o 














> 






o 


c 


a 


■2 


o 


o 


ni 


to 


to 


to 


nj 


"3 „ 


nj 



en 

00 
ON 





10 




CO 








cd 


CM NO 


O O O CN) 


O 


00 


VD -H; 


00 


NO 


J" 


CM 


- - 

NO 00 




C CO 


rH 




cu 


cn r^ 


rH O O O 


O 


UN 


CO IN- 


-a- 


CM 


O 


rH 


O <~> 




•H 0) 


-P 




E 


CNl 








rH 


CM 














faO-H 


c 






























•H CJ 

^1 




CJ 




J- 


Cn- UN 


CN) O O rH 


O 


O 


NO CO 


ON 


CM 


O 


CN) 


r-l CI 




O ft 






rH 


CN 






rH 




r— 1 












• 


co 


c 


1 — 1 

M ° 




3 
























5 


0) M 

x: id 


o 

•H 




^^ 
























3 


Eh Xj 

H 3 * 


•P 

3 
X) 




PQ C^ 

r5 ^ 


c^ ° 


OOOO 


O 


CM 


J- (N- 


rA 


O 


O 


O 


O O 


g> 


cd -a 

CD T) 3 


•H 
rH 


CO 




O 


























-a • -p 


-P 


X. 




Eh CN1 


CN] O 


rH O O O 


O 


rH 


rH J" 


O 


rH 


CM 


CN 


O -H- 


a 


•h cd co 

s n) w 


CO 

■ H 


s$ 




M rH 
Eh 


t-i CM 






rH 


rH rH 


UN 












»M 


h H 
e ni x: 3 


•d 


o 
-p 




Ph 
























s 


Eh O 


M 


o 3 




rH 


rH VO 


CN) O O O 


O 


CN 


CV CN 


CM 





O 


O 


O 


o 


CM >> 


o 




w r-i 


rH 








CM 


rA 














d • C 


ch 


■p 




CO 
























§ 


CO 3 ft -H 




CO 




























•H +J ft 


-d 


O 




< n 
























« 


CO CO "d 





-P -rH 







CN) C^- 


OOOO 


O 


CN 


UN UN 


ON 





O 


r-i 


CNJ 


o 


-P 


CO 


-p 




rH 


C\) 








H CN 


CM 


rH 










co 


0£ 10 


-p 


3 


-H 




























Q 


a) o 3 
co nJ C 
c cd 
cd x) 


o 
c 

to 


CO X) 

-32 


l-i rH 
X) -H 

™ id CO 
cd -P -p 




























^ 
























M 


h c c 


a 


XI O 




S c 
3 cd 


O CN 


CN 00 O CN 


CN 


CN 


UN u^ 


CN 


O 


O 


O 


CV c-\ 


£ 


■"3* 


u 


E CO 


rH Cd (rt 
Cd rH O 

ftcd hD 































>> 




Ph 


ON ON 


CN) CN) CN NO 


CN 


CN 


CN- NO 


MD 


J-" 


CV 


CN 


CM r-» 


55 


x: -p -p 


co cd 




O E 


CN 


rH 




rH 


rH -H- 


UN 




CM 


rH 






M 


O -H CO -P 


rH 


rH 




























d 4 3 jl 

d) rH ft 


3 
x: -d 


II u 
o 


w 


























co 


























Eh 


CO ft 


-P -rA 


Eh • <M 


:t; 


























O 


C 


o 


^ 


\j 


























Cd 


« rH --P 


Id 


. rH • 





G 


O O 


^OOO 


UN 


UN 


O O 


UN 


UN 


O 


UN, 


UN 




co cd 




m «> ra 


H 


cd 
























Z 


rH -p -p h 


^^ • 


£ > rQ 

y - h o 

Co ^J 

-d o • 


Cx. 





rH O 


OOOO 


CN) 


NO 


O O 


NO 


CM 


CN) 


CM 


i-H r-» 


S3 

En 


xi -p o -P 


CN Ph 




ixl E 


CN) rH 








r-1 CM 


H)- 












id cd CO 
-P ft co 3 


00 
ON II 


I 


























S5 


C rH 

co g a h 

•H O H -H 


H P, 


CO 
CO ft U 


cn 


Ph 

O ON 


C^ NO 


OOOO 


UN 


O 
rA 


00 CN 

rH 


C^ 


CN 


O 


CM 


O O 


O 


c 


3 CO 




























M 


XI -rH -p 


o 


ft 




• 
























En 


-P -P o 


CO •• 


CO U 


S 


Cd 


"N. J" 


rH O O O 





CN 


CM CN- 


ON 


CN) 


.j- 


CN 


CN 


Ih 

Cd 


cd -P 

G -P CO 

•rH (1) >d 


ftd 

• H 
rH rH 


rH Tj 
O - C 

xi co id 


CO 


CN 








1 ■ 


UN 












S 




























hox: 


rH -P 


E -P 




























> 


d CD -P u 
> 


•H G 

x: 


tiOOO 


" 


C 
cd 


O >n 


O U^l O "N, 





O 


O O 


O 


UN 


O 


UN 


O UN 




co C -d 


Ch O 


H 





Dn m 


4t ICN NO CN) 


-3- 


O 


UN CN 


NO 


UN 


CM 


CN 


CN f-» 


s 


3 O -H .rH 




ON 


CJ 


:*: e 


CN) rH 




CM 


cn) r^ 


NO 




UN 


CM 






M 


-P CO 


•d -P 


XI -d rH 


a 


5 


























CU T) C 


C CO 


H-> C v_ ■■ 


m 
























CO 


^ -p g o 


cd cd 


cti • 


S3 


Ph 


CO O 


J- CN CO CN- 


-d- 


CN) 


CM ON 


ON 


NO 


rH 


CN 


UN 


BQ 


cd O 3 O 





rj 


1 


r^ 


»A 






rH 


CM O 


CN 




CM 


rH 






M 


cu o 


c x: 


— co cdl 












r-i 














O 


rH ft Ch CU 


O -P 


CO 


H 


• 
























I 


CO U 

cu cu co nJ 


CO 3 
XI o 


CO -PI 

H -rH 


a 


NO 


\0 c^- 

"N rH 


J- 00 4; 00 


J- 


00 
CN) 


00 (N- 
<N) CN 


CN 
ON 


UN 


CN 
00 


C^ 


rH O 


CO 


U U -P 
3 C -p 


•H CO 





K 




























o 


> -P M) 

•H H C 


•: 


























| 


box: nj 3 


II 




























•ri-PH^l 


u 


X) O -H 




























<J 


Ph -H ft 


Cd 


M -P S 


u. 


c 


VO NO 


O Hf O 


NO 


NO 


NO CN) 


00 


00 


O 


J- 


CM cv) 




s 


-P CO 







cd 
























P-l 


C >->-P 


cm 


x: z 


< 





00 CN) 


00 00 MO O- 


r^ 


CN 


UN, CM 


CV 


00 


CN- 


C*. 


_7 


^H 


Q 


•h e -b c 

O O CU 


cd 


■p 



>H 


E 




"N rH 




r-t 


rH CM 


-3- 




ct 








g 


c tj o e 


CO 


so >>^ 




























£ G S -P 


C 


rH -P CN) 


^J 




NO ON 


rH O CN- UN, 


rH 


C<t 


O rH 


3 


CN 


t-i 


j- 


^ 


O 


O 


o d cd 

x: U ch cu 


o > 


cd -h 00 


In 


"N 


rH 


CN) CN) rH 




CM 


CM CM 


t-i 


J- 






9 


•H O 


rH rH ON 


cn 


























co o h 


•P H 


•H rH 




























Cd 


-P -P 


3 t>D 


O XI-—- 






CN CN 


NO NO J" C^ 


ON 


rH 


rH 00 


r-t 


J- 


3 


O 


UN 1 


CO 


co id to 

-P rH 


rO 

•H || 


^HHC 




J- 


rH 




rH 


rH rH 


J- 


rH 


rH 






Ch 


o -H n) 


rH 


co ni 




C-H 
























O 


■P O 


rH CO 




pd 


CO O 


-3- CN rH CM 


UN 


O 


CM UN 


r-{ 


rH 


UN. 


00 


.-J 


r^ 


cn 


CO -H 
G O ft-P 


CO 

-H 


•H cd C 
•p ft 




g CN 




IN- CN) 




<-\ 


CM CM 


J" 


rH 








>-H 


cd -P co to 


•d 


•rH C > 




O 
























< 


M -H 




rH 3 




rH 


NO rH 


CM rH 00 NO 


^1- 


M3 


J" [N- 


UN 


CN) 


CN 


CN 


C J 


rH 


s 


■Ptl c^ 


Cm -d 


•H +J 




O CNJ 




J" rH 






CN) 


1— 1 




CM 






o 


o cd 

rH E -P 


O 


X) CO 




























H 


M 


Cd -P 
-P O -d 




























> 


£1 E CO 


C -p 






O O 


CV CM O u>i 


ON 


O 


rH O 


d- 


.j- 


NO 


CN) 


O 


CNJ 


M 


-p d o 

•H O rH 
CH CO O 


o c 

•H 
-P o 


cd C C 

rH Cd 




rH 




CN) CO CM 


rH 


rH 


CN) CM 







C^- 


rH 






£2 


























l-H 


O G -P <H 


o 


























Eh 


O CO 

e o o -p 


•H HJ 
U CO 


U CO 

c 3 


$ 


y 


OOOO 


O 


O 


u 


O 


O 


5 


s 


O 


1 


O 

EC 


E G 

O 

O h^'ri 


-p 

CO 5 

x: 


S co 

-d -h 
ceo 


Ph 


^ B 


Eh* H 

3 


B 


i 


£ £ 


Eh 

3 


s 


B 


b" 


^ 


Cd 


rH flj rH O 
•H 


rH +j 

H 


.* cd -p 

rH 


























. 






















1 


5? 


•P CO <H 


CO o 


CO Eh O 
O -P 


Eh 





CJ 























CO -P x: =M 
o cj -p 3 
E co 
2 co C 


c 


co 
























S 


-p 

O II 

a 


-P 

O H 
C - O 


H 

n 


3 Pd 

S CO 


% § 












1 








I 


















*/} 










m G H -H 


s 


CO Ch 


















■H 










Eh 


IB Ji id 

x: u >> 


•d S 


TJ "P 

tjDIN- 




co 


cd 






CO 








o5 




H 








g 


-P -P TJ rH 
W rH 


. 





•d 




B 


co 


H 

5 


| 





■ H 

cn 




1 


cd 


(d 




rH 

r— | 




•H 






Pi 


M 


O -P nj 

c x: o K 


• 


ON 


CO 


.H 


i 




B 


>> 


cd 


rH 


(0 




C 


c 





•H 


rH 


c 




•H 




-p 






P« 


* 


Eh U 


• O rH 


a 










£ 


P 


• H 


rH 


c 




a 





•H 


u 


O 







X 




§ 


rt 


-iH 


to 


o 


-p 


W Cd 


1^ 


1— 1 


rH 






-p 





CU 


-i * 


<ti 


■"I 


>- 







H 


rH 


=H 


cd 





H 


rH 




fd 




■r- 


• H 




BC 


® rH C 


H 





| 


:>, 


CO 


• rl 







$ 





a 




• H 


x: 


H 


CO 


rt 


>i 


O 




nj 


Ch 


^ 


<U 




i 







H 


c 


w 


CO 


x; <m 


o 


Ph 


a 


X 





■d 


2 


cd 





CD 


X^ 


ch 


P 


XI 


g 


of 


X 


■H 


CO 


3 


2 


•r- 


X 


E 


cti 


g 


H 





a 


■H 


3 




Eh O to hO 


» 


■ 


K 


P 


1 c 


-p 


•H 


• H 


U 


co 


u 





H 


H 


■H 


O 


5 


c 


r - 


O 


x: 


• H 


■n 


P 


-P 





M 


j^ 


g 


E 


> 


c 


i 








m 


> 


, a 


>! 


E 


x: 


XI 








cd 


Cm 


O 


| 


•P 


^ 


c 


H 


HP 


h 


•H 





rH 


H 


Cb 


rj 


c H 


H 


• H 


• H 


cd 










r-i 


1-1 




• H 


+- 1 


•H H 


.^ 





V 


x: 




rH 


p 





HJ 





Ch 


CJ 


>M 


3 





c 


CD 


3 


m 





E 


O 


TJ 


Eh 












cd 


< 




< 


■dl fj 

cd 
O 


Xll O 

• H 
O 


E g 


H ft 
3 
Cd 


p5 e 


H 


cd w 


co co 


«J E 


Ph 


Cd rH 
O 
Ph 


E d 
•H 
CO 


• H 


H 

nj 

Eh 


03 ii 

Eh 


> rH 



Eh 


r^ 


e 

Ph 



ATOLL RESEARCH BULLETIN 
No- 291 



CHANGES IN THE DISTRIBUTION OF THE COCCID 

ICERYA SEYCHELLARUM WESTW. ON ALDABRA ATOLL 

IN RELATION TO VEGETATION DENSITY 



By 
D. McC- Newbery and M. G- Hill 



Issued By 

THE SMITHSONIAN INSTITUTION 
Washington, D. C«, U.S-A- 
May 1985 




00 


4-1 


c a 


■H - -H 


u e 


o 


O 3 


a 


4-1 >- 




•H CO 


00 


a 4J 


c 


O 3 


'H 


e to 


tH 


c 


a 


14-1 


o co 


cfl 


3 


CO 


CO o 




C -H 


14-1 


O Pn 


o 


•H 


4-1 - (-1 


HZ 11 


U fn J3 


O E 


■H -.3 


cfl 


c 


CD -H 




x: -h 


<d 


•u o 


X! 


U-4 


4-1 


T3 -H 




c u 


o 


CO >i 


4-1 


a 




^ M 


0) CO 


CD 


> -H 


<4-l 


U X> 


<d 


3 M 


u 


CO O 




a) x 


CO 


m a 


u 


3 


0) 


m w 


X 


oo E 


0> -3 


<H P-, 2 


W 


OJ 


X « • 


4-1 CO 


CO 




c 


— 1 




C -H 


CD 




•H »-l 


o 




CO 


o 




T3 6 


CO 




0) 4J 




i-H CO 




D. -rl 


CO 




E C 


rH 




CO C 


o 




en ai 


> 




o 


"J 




CD -H 


CD 




CO > 


CJ 




CD < 


Cfl 




)-i 


CO * " 


2 H 
01 < CO 


43 


4-1 •> 


.. a; 




oo co e 




C CD t-i 




•H -H OJ 




3 O C 




O 0) -H 




x o. 


CO CO c 




o 




rH CD .-1 




.-H X >i 




O 4J X 




4-> O 




< u u 




O CD 




CO <4-l TD 




l-l -H 




£ CO CO 




CO 01 







< CO CO 



00 
•H 
Pn 



CHANGES IN THE DISTRIBUTION OF THE COCCID 

ICERYA SEYCHELLARUN WESTW. ON ALDABRA ATOLL 

IN RELATION TO VEGETATION DENSITY 

By 
D- McC Newbery^nd M. G- Hill 2 

ABSTRACT 

1. The overall abundance of Icerya seychellarum Westw. (Margaro- 
didae : Homoptera) , on Aldabra Atoll in the Western Indian Ocean, has 
changed little between 1978 and 1983 but its spatial distribution over 
the atoll has altered markedly. 2. Several susceptible host tree spe- 
cies showed ten-fold or more higher median infestation levels in the 
SE than the NW quadrant of the atoll. These differences were not evi- 
dent in 1978. 3. The results of the 1983 survey are supported by 
biannual monitoring of five species of host trees from 1980 to 1983. 
4. Since 1978 the level of coccid infestation has risen in the SE, 
where tree mortality is largely density independent, but has remained 
low in the NW, where tree mortality is more density dependent. 

Keywords: Coccids Aldabra Atoll Distribution Hosts Vegetation 
Density. 

INTRODUCTION 

Two surveys of the distribution of the coccid, Icerya seychellarum , 
Westw. (Margarodidae : Homoptera), in 1976/77 and in 1978 on Aldabra 
Atoll, showed an overall low abundance (Hill and Newbery, 1980) com- 
pared with the peak infestations of 1975 (Renvoize, 1975). Host tree 
species differed considerably in their susceptibility to coccids (Hill 
and Newbery, 1980; Newbery, Hill and Waterman, 1983) but the abundance 
of coccids on any host species varied insignificantly between different 
areas of the atoll. Thus, by late 1978, the outbreak which had started 
c. 1968 (Hill and Newbery, 1982) appeared to have settled down to an 
even and residual level. In this paper we report the results of a re- 
survey of the status of I_. seychellarum on Aldabra in 1983 and monitor- 
ing of the coccid on five tree species between 1980 and 1983. 



Department of Biological Science, University of Stirling, Stirling 
2 FK9 4LA, Scotland. 
D.S.I.R. Entomology Division, Private Bag, Auckland, New Zealand. 



METHODS 

Survey of coccid distribution and abundance 

Between July and August 1983, we recorded the abundance of I_. 
seychellarum on the woody flora of Aldabra in the areas shown in 
Figure 1. Apart from the recently-opened long transect in the south- 
west, these areas were approximately the same as those sampled in the 
1976/77 and 1978 surveys by Hill and Newbery (1980). 

The woodlands and shrub vegetation were sampled by a plotless 
method. In each area sampling points were chosen either in a pseudo- 
random, representative manner, or evenly at intervals of c. 100m along 
transect lines. At each sampling point the nearest thirty trees were 
identified and the coccid infestation scored as a composite value for 
each individual tree and on the same scale used in 1976/77 and 1978, 
viz: 0, no coccids found after searching; 1, few coccids found after 
searching; 2, numerous coccids evident without searching; 3, vast num- 
bers of coccids; 4, coccid infestation devastating. Our hundred and 
twenty-three sampling points were all in 'mixed-scrub' vegetation 
(Newbery and Hill, 1981; Gibson and Phillipson, 1983) and these were 
supplemented by observations on all individuals of susceptible species 
when encountered in walking between sampling points. In this context 
a susceptible species was one which had a median infestation score of 
1.0 or more in the 1976/77 or 1978 surveys. Mangrove, Pemphis acidula- 
dominated, and coastal scrub communities were sampled in a similar but 
less intensive manner. (Authorities for plant species are included in 
Table 1.) Time was not available to comprehensively search for rare 
tree species. 

The results of the late 1983 monitoring (referred to below) were 
added to our sample. Excluding seven uncommon host-tree species, each 
of which had less than ten individuals, 5137 trees and woody shrubs 
were sampled from thirty-six species. 

Monitoring 

In the light of the earlier surveys (Hill and Newbery, 1980) and 
detailed studies on highly susceptible species (Newbery, 1980a, b,c 
and Hill, 1980), thirty individuals of each of five heavily infested 
tree species were randomly selected and tagged along transects in areas 
of the atoll where they were locally abundant and highly infested in 
1980 : Scaevola sericea (near the Research Station, Picard Island, and 
near Dune Jean Louis, Grande Terre) ; Euphorbia pyrifolia (Anse Var, 
Picard Island); Ficus nautarum (Cinq Cases, Grande Terre); Sideroxylon 
inerme (Gionnet and Middle Camps, Malabar Island) and Avicennia marina 
(La Gigi, Picard Island, and Cinq Cases creek, Grande Terre). 

Trees were scored for coccid abundance biannually from 1980 to 
1983 using the survey scoring scale of to 4. Where an individual 
tree died between the sampling times the number of replicates were 
fewer at the time of measurement and these numbers were made up by 



randomly selecting and tagging new individuals. (_E. pyrifolia was not 
sampled in the first six months of 1980: In one case, S^. sericea near 
Dune Jean Louis, the number of replicate bushes was twenty-three; and 
in twelve other of the sixty-four species-site-time combinations this 
number was between twenty-six and twenty-nine.) 

RESULTS 

Comparison with previous surveys 

The median infestation scores of the thirty-six species are shown 
in Table 1 for the whole atoll and for its four quadrants (Fig. 1). 
Thirty-two species had median infestation scores greater than nil in 
either the 1983 survey or one of the 1976/77 or 1978 surveys (Table 1, 
Hill and Newbery, 1980). Spearman rank correlations between the scores 
in 1983 and scores in 1976/77, and in 1978, were highly significant 
(P <0.01, r=0.773 and 0.864 respectively), indicating that those spe- 
cies which were highly and lowly infested in 1976/77 and 1978 remained 
so in 1983. 

For the whole atoll the changes in median infestation for these 
thirty-two species between 1976/77 and 1983 and between 1978 and 1983 
were not significant by sign tests (15+, 17- ; 18+, 10- , 4 nil differ- 
ences respectively, P >0.05). Of the ten most heavily infested species 
in 1983 (Table 1), between 1976/77 and 1983 five increased and five de- 
creased in infestation and the same occurred between 1978 and 1983. 

Variation in infestation across the atoll (1983) 

Ficus nautarum , Sideroxylon inerme and Apodytes dimidiata show 
approximately ten-fold higher mean infestation scores in the SE quadrant 
than in the NW quadrant, and Avicennia marina shows a sixteen-fold dif- 
ference in the same direction. For Pemphis acidula the infestation in 
the SE is about sixty times that in the NW, though it must be noted 
that this is coastal and mixed-scrub _P. acidula and not the main P_. 
acidula dominated vegetation zone in which individuals are rarely in- 
fested. Scaevola sericea and Polysph aeria multiflora differed little 
in infestation between the SE and NW quadrants : many other species in 
Table 1 had either sample sizes that were too small for analysis or 
median infestations that were too low to make comparisons across the 
atoll. The NE, like the NW, quadrant had comparatively low infestation 
scores, even on the generally more susceptible species; but in the SW 
quadrant (most of which was not accessible for sampling in 1976/77 and 
1978) Sideroxylon inerme , Scaevola sericea and Polysphaeria multiflora 
had notably high and similar levels of infestation to those the SE 
quadrant. 

Clearly, in the SE quadrant the susceptible host species support 
higher median infestation levels than in the NW quadrant with the SW 
and NE quadrants being intermediate. Comparing the median infestation 
scores in the NW quadrant (1983) with those over the whole atoll in 



1976/77, and in 1978, shows no significant change (P >0.05) by sign 
tests (8+, 9-, 2 nil ; 7+, 11-, 1 nil differences, respectively). 
Similarily, for the SE quadrant the sign tests show an overall rise, 
but this is significant at only P $0.1 (10+, 4-, 1 nil differences in 
both cases) . 

Differences in weig hted a bundance of coccids across the atoll 

In 1976/77 Newbery and Hill (1981) recorded the percentage cover 
abundance of trees and woody shrubs in sixty-five, mostly 20 x 20 m 
plots, during the course of that coccid survey. The 1983 coccid data 
were collected from the same areas (comparing Fig. 1 in Hill and New- 
bery, 1980 with Fig. 1 here) and there has been little visible change 
in the vegetation and its composition in that time interval (excepting 
unusually damaged vegetation inland of Dune Jean Louis - Newbery per- 
sonal observation 1983). Those vegetation data may be used to calcu- 
late weighted vegetation coccid scores for the 'mixed-scrub' areas in 
the NW and SE quadrants. The NE was less intensively sampled for 
coccids, both earlier and in 1983, and the flora of the SW has, in its 
widest part, only been recently investigated (C. Peet and D. Cowx 
unpublished) and not with comparable plot records. The thirty-three 
commonest species were used : fourteen rarer (less than 1% cover) and 
very infrequently infested species were excluded. The mean percentage 
cover abundance for each species was calculated for thirteen NW quad- 
rant plots and twenty-eight SE quadrant plots. (Of these latter, one 
was of coastal Scaevola sericea dominated scrub and two others lay in 
the Thepesia populneoides - Lumnitzera racemosa association inland of 
Cinq Cases creek) . The coccid score for each host species in the two 
quadrants was weighted by the hosts' mean cover abundance and the 
overall weighted mean found. For the NW quadrant this mean (still on 
the to 4 coccid scoring scale) was 0.120 and in the SE quadrant was 
0.363 - a three-fold difference across the atoll. 

Changes in infestation over four years 

Coccid infestations on monitored tree species generally continued 
to decline on Aldabra between 1980 and 1983, except for the SE moni- 
tored Avicennia marina and Ficus nautarum (Fig. 2) which appeared to 
increase. Infestations have been high on Scaevola sericea , Avicennia 
marina and Ficus nautarum to a similar level shown in the 1976/77, 
1978 and 1983 surveys. The most obvious decline from moderate infes- 
tations in 1980 to near zero levels in 1983 were for Sideroxylon inerme 
(Malabar Island) and Avicennia marina (Picard Island). 

DISCUSSION 

The two previous surveys of 1976/77 and 1978 were conducted in 
Aldabra 's wet season whilst access to the atoll was only possible at 
the end of the wet season and the start of the dry season in 1983. 
It is unlikely that differences between the 1983 and previous results 
were due to meteorological changes because, apart from species like 



Euphorbia pyrifolia which are deciduous (Newbery, 1980b), phenology 
was not a significant factor in host tree susceptibility (Newbery, 
Hill and Waterman, 1983) and monitoring over the four years (Fig. 2) 
did not show periodic changes in infestation levels from dry to wet 
season. Overall, there has been little qualitative change (in the 
terms of numbers of host species increasing and decreasing) in the 
status of I. seychellarum on Aldabra between 1976 and 1983, and this 
is supported by the monitoring results. Within the atoll there have 
developed marked differences between the NW and SE quadrants with in- 
dications of serious local increases in the SE between 1976 and 1983. 

Aldabra has become drier in recent years. Stoddart (1984) has 
analysed eight years of atoll rainfall patterns based on a rain-gauge 
circuit of thirteen stations around the atoll (1974-1981). Grouping 
the yearly total rainfall results for stations within each of the four 
quadrants (and excluding that for lie Esprit) for the years 1976 to 
1981, the NW quadrant decreased by 41% from 1695 mm to 999 mm. Simi- 
larily, for the NE, SE and SW quadrants respectively the changes were: 
1209 to 798 (34%); 1229 to 856 (30%); and 1273 to 855 (33%). The SE 
is consistently drier than the NW, though suffered a slightly smaller 
decrease in rainfall than the NW between 1976 and 1981. 

It seems unlikely that the infestation in the NW was lower because 
an agent of biological control has taken effect in these recent years. 
In 1983 no coccinellid beetles were seen and in 1976-1978 Hill and 
Blackmore (1980) found only a few beetles after searching. Parasites 
were not found in 1976-1978 and in 1983 there was no evidence of dead 
colonies which could have been a result of these or of a pathogen. 

There are large differences in habitat between the NW and SE 
quadrants of the atoll. In geomorphology , the NW has a rough, dis- 
sected terrain of pave and champignon coral whilst in the SE the 
predominant form is flat platin limestone (Stoddart e_t al_. , 1971) 
except for the areas just inland of the south coast. As a consequence 
soils fill wide shallow basins in the SE whereas in the NE trees are 
rooted in much smaller, often deeper pockets (Trudgill, 1979a,b). The 
vegetation in the NW is a species-rich closed canopy (average percent- 
age cover of woody species 131%, Newbery and Hill, 1981) compared with 
the relatively species-poor and more open (85% cover) canopy in the SW 
(see also Hnatiuk and Merton, 1979a, b and Gibson and Phillipson, 1983). 
The vegetation in the SE is more exposed than in the NW to SE trade 
winds during the dry season (Hnatiuk, 1979). These factors suggest 
that decreased rainfall will be more deleterious to the vegetation in 
the SE than in the NW, and greater water stress may, in part, explain 
the higher levels of coccid investation in the SE (Newbery, 1980a, c). 

Against this hypothesis is the observation that the mangrove 
Avicennia marina will not be short of water in dry season yet this 
species shows one of the greatest differences in infestation between 
the NW and SE quadrants. Newbery (1980a) has suggested that one of 
the controlling factors of infestation on A. marina is the build-up 
of excreted salt on the younger leaves and therefore frequent rain 



may keep leaves more receptive to coccid settlement. Possibly in- 
creased immigration onto A. marina from other stressed plants led to 
high levels in the SE. 

An alternative, but not isolated, hypothesis follows from Hill 
and Newbery (1980). The peak infestation levels in 1975 (Renvoize, 
1975) were at levels far greater than those we recorded in 1976/77 and 
1978 and may have caused the death of some susceptible trees (Newbery 
1980b, c). These deaths would have thinned the vegetation and left 
young, more resilient, individuals. Could this have happened faster 
in the NW than in the SE? In the SE the trees are more widely spaced 
and mortality is probably density independent in the main due to en- 
vironmental factors (Stoddart and Wright, 1967) and to grazing (feral 
goats, Gould, 1979; and tortoises, Merton, Bourn and Hnatiuk, 1976) 
which predominates in the SE. In contrast, tree mortality in the denser 
luxuriant vegetation of the NW is likely to be more density dependent. 

In the NW there is ample evidence of tree regeneration (Newbery 
pers. obs.), whereas in the SE the grazers reduce seedling survival and 
hence regeneration, especially in the areas where grazers and coccids 
are both abundant. Removal of phloem sap by coccids in dense vegeta- 
tion will mean that an infested, and therefore weakened, tree (Newbery 
1980b, c) is less able to compete with its non-infested neighbours and 
would be rapidly thinned from the vegetation. Where there is sufficient 
rainfall, trees could maintain a rapid leaf turnover rate leading to 
high rates of leaf mortality for a sedentary stylet feeder such as 1. 
seychellarum , (Hill, 1980). Conversely, in the SE the same species, 
less affected by tree-tree competition, although debilitated by coccids, 
would be expected to survive longer and to either have a slower leaf 
turnover rate or become deciduous as a result of the drier environment. 

Evidence from several species supports this role of vegetation 
density in the population regulation of I. seychellarum : 

1. The highly infested fig trees (Ficus nautarum ) in the SE are large, 
imposing trees whilst in the NW they tend to be smaller, growing in 
amongst other shrub and tree species. Similarily, the heavily infested 
Guettarda speciosa trees in the SE are well separated from neighbours 
and therefore probably suffer less competition as a result. Apodytes 
dimidiata , commonly infested in the SE, does grow in small clumps of 
trees and shrubs though not infrequently as separated individuals. 
Lastly, Avicennia marina stands sampled at La Gigi, Picard Island (New- 
bery, 1980a) and at Cinq Case creek are structurally different : The 

NW site shows colonization on a sand bar with young growth and competi- 
tion, whereas the trees in the SE are larger and more spaced in coral 
pockets at the upper limit of the tide and where the zone of brackish 
pools begins. 

2. Species such as Scaevola sericea that tend to grow as monospecific 
stands of similar density in the NW and SE, suffer intraspecif ic com- 
petition between similarly infested individuals and showed little dif- 
ference in infestation levels between these quadrants (Table 1). 



Polysphaeria multiflora is a small tree species of dense mixed-scrub 
and woodland over most of the atoll (Newbery and Hill, 1981; Gibson 
and Phillipson, 1983) and, rarely being found as an isolated individual, 
also showed similar infestation levels in the NW and SE. 

3. Lumnitzera racemosa and Thespesia populneoides do not afford a NW - 
SE comparison as they form a special community type only in the SE. 
Calophyllum inophyllum dominates an isolated grove in the SE, and 
Casuarina equisetifolia occurs in the NW and NE. However, for T. 
populneoides (much less so.L, racemosa which lines brackish pools), 

it was common to find well spaced individuals which were frequently 
heavily infested (Table 1). 

4. Pemphis acidula mainly occurs as an almost monospecific band of 
vegetation around the atoll inland of the lagoon mangroves (Gibson and 
Phillipson, 1983), and there it is very lightly infested. In the 
sparser vegetation along the SE coast it had moderate infestation 
levels. Sideroxylon inerme provides an interesting case, because it 
also had higher infestation levels in the SE than in the NW. In the 
NW this species is commonly found in either dense P_. acidula stands 

or in mixed-scrub, whereas in the SE it occurs as isolated trees. 

Our findings and hypothesis illustrate an important ecological 
principle. This new immigrant insect to the island ecosystem of 
Aldabra is still settling into its fundamental niche (MacArthur & 
Wilson, 1967; Pianka, 1978). The extent to which this niche is de- 
veloped differs in different vegetation types. In the parts of the 
atoll (NW) where the vegetation appears to be near equilibrium we 
suggest that the original outbreak has been dampened to a residual 
level by a process of negative feedback (thinning and the capacity 
for vegetation regrowth) whereas in the non-equilibrium parts (SE) 
which are subject to stronger environmental stresses enforcing the 
effects of increasing large herbivore pressures on the vegetation, 
a recent positive feedback has occurred in the form of a small up- 
ward oscillation in coccid abundance. On this basis, we predict that 
coccid abundance will fall back to a residual level in the SE once 
the susceptible host trees have all died out. And, since the regen- 
eration of these host species is limited in the SE, this residual 
level may well be lower than that in NW, not precluding the possi- 
bility that in a few decades present young individuals of susceptible 
species will have aged and become more infested in the NW. 

ACKNOWLEDGMENT 

We thank past wardens of Aldabra, C. Peet , J. Stevenson, and R. 
Pimm for recording monitoring data; the Carnegie Trust for the Uni- 
versities of Scotland for travel funds (for D.McC.N); the Royal Society 
and the Seychelles Islands Foundation for permission to revisit the 
atoll; and D. R. Stoddart and L. U. Mole for encouragement in all 
matters Aldabran. 



REFERENCES 

Gibson, C. W. D. & Phillipson, J. (1983). The vegetation of Aldabra Atoll : 
preliminary analysis and explanation of the vegetation map. Philosophical 
Transactions of the Royal Society of London, j$, 302 , 201-235. 

Gould, M. S. (1979). The Behaviour ecology of the feral goats of Aldabra 
Island. PhD Thesis. Department of Zoology, Duke University. 

Hill, M. G. (1980). Susceptibility of Scaevola taccada (Gaertn.) Roxb. 
bushes to attack by the coccid Icerya seychellarum . Westwood: the 
effects of leaf loss. Ecological Entomology, _5_, 345-352. 

Hill, M. G. & Blackmore, P. J. M. (1980). Interactions between ants and the 
coccid Icerya seychellarum on Aldabra Atoll. Oecologia (Berlin), 45 , 
360-365. 

Hill, M. G. & Newbery, D. McC. (1980). The distribution and abundance of the 
coccid Icerya seychellarum . Westw. on Aldabra Atoll. Ecological 
Entomology, _5, 115-122. 

Hill, M. G. & Newbery, D. McC. (1982). An analysis of the origins and 
affinities of the coccid fauna (Coccoidea : Homoptera) of Western Indian 
Ocean islands with special reference to Aldabra Atoll. Journal of 
Biogeography, £, 223-229. 

Hnatiuk, R. J. (1979). Temporal and spatial variations in precipitation on 
Aldabra. Philosophical Transactions of the Royal Society of London, B^, 
286 , 25-34. 

Hnatiuk, R. J. & Merton, L. F. H. (1979). A perspective of the vegetation of 
Aldabra. Philosophical Transactions of the Royal Society of London, J3_, 
286 , 79-84. 

Hnatiuk, R. J. & Merton, L. F. H. (1979). The vegetation of Aldabra : a 
reassessment. Atoll Research Bulletin, 239 : 1-22. 

MacArthur, R.H. & Wilson, E.0. (1967) The Theory of Island Biogeography . 
Princeton University Press, New Jersey. 

Merton, L. F. H., Bourn, D. M. & Hnatiuk, R. J. (1976). Giant tortoise and 
vegetation interactions on Aldabra Atoll - Part 1 : Inland. Biological 
Conservation, 9_, 293-304. 

Newbery, D. McC. (1980a). Infestation of one coccid, Icerya seychellarum 
(Westw.), on the mangrove Avicennia marina (Forsk.) Vierh. on Aldabra Atoll, 
with special reference to tree age. Oecologia (Berlin), 45^, 325-330. 

Newbery, D. McC. (1980b). Interactions between the coccid, Icerya 

seychellarum (Westw.), and its host tree species on Aldabra Atoll. I. 
Euphorbia pyrifolia Lam. Oecologia (Berlin), 4_6, 171-179. 

Newbery, D. McC. (1980c). Interactions between the coccid, Icerya 

seychellarum (Westw.), and its host tree species on Aldabra Atoll. II. 



Scaevola taccada (Gaertn.) Roxb. Oecologia (Berlin), 4_6, 180-185. 

Newbery, D. McC. & Hill, M. G. (1981). Numerical classification of 
'mixed-scrub' vegetation on Aldabra Atoll. Atoll Research Bulletin, 
246:1-13. 

Newbery, D. McC, Hill, M. G. & Waterman, P. G. (1983). Host-tree 

susceptibility to the coccid Icerya seychellarum Westw. (Margarodidae : 
Homoptera) on Aldabra Atoll : the role of leaf morphology, chemistry and 
phenology. Oecologia (Berlin), 60_, 333-339. 

Pianka, E. R. (1978). Evolutionary Ecology . 2nd ed. Harper & Row, New York. 

Renvoize, S. A. (1975). Icerya seychellarum on Aldabra. Unpublished Royal 
Society Aldabra Research Station Report, ALD/21(75), 4lpp. 

Stoddart, D. R. (1984). Spatial and temporal variability of rainfall on 
Aldabra Atoll. Atoll Research Bulletin (in press). 

Stoddart, D. R., Taylor, J. D., Fosberg, F. R. & Farrow, G. E. (1971). 

Geomorphology of Aldabra Atoll. Philosophical Transactions of the Royal 
Society of London, B_, 260 , 31-65. 

Stoddart, D. R. & Wright, C. H. (1967). Geography and ecology of Aldabra 
atoll. Atoll Research Bulletin, 118 , 11-52. 

Trudgill, S. T. (1979a). Surface lowering and landform evolution on 
Aldabra. Philosophical Transactions of the Royal Society of London, B_, 
286 , 35-45. 

Trudgill, S. T. (1979b). The soils of Aldabra. Philosophical Transactions 
of the Royal Society of London, B, 286, 67-77. 



10 



c-> 
oo 



►4 
W 

o 



oi ooofflin on o <r 
m f^ * oo iri ^d o\ Or- 1 

rH .H -Ct CN 



iH "1 "1 viO CI CO O -* O 

>*oO^OOOOo> r-^r^ 



OOMNiOlMH-JMSHHOO Ol 

00\Of r )»T>*lO<rNO\OOClT-( 00 

iH oo ci en 



ooO-Joooo*Dinrs^-jooD* ^ 
^m^cicinnMNriOO O 



CN 

o 



rH iH iH iH O 



O O 



OOOOOOOOOOOO 



a 

3 

u 

■H 

u 

X 
cj 

OJ 

J) 



UH 




o 




n 




OJ 




u 




o 




o 




CO 




c 









•r-l 


*r""s 


u 


OJ 


eg 


N 


4J 


■H 


CO 


DO 


0) 




M-l 


OJ 


C 


rH 


•H 


C 




e 


c 


id 


co 


CO 


■H 




-o 


oj 


OJ 


JB 


s 


4J 



XI 
OJ 

H 



en moo o> n O n 

U-| v£) LO LO rH 



o 








<-l 


in 


o 


iH 






o 


o o 


4J 








<t 


oo 


o 


vO 


1 


1 


m 


(Ti r-» 


< 




w 




• 


• 




• 


1 


1 


• 


• • 






LO 




CN 


H 


ro 


H 






rsl 


o o 


(0 
























U 
























XI 
























CO 
























-a 








o 


O 


ON 


O 


o 


o 


O 


m o 


H 






c 






rH 






>o 


.H 


CM 


< 




















iH 




<4-l 
























o 












^O 






CN 


<r 


<t 


CO 








1 


1 


O 


1 


1 


CJ> 


m 


O 1 


M 




w 




1 


1 


• 


1 


1 


• 


• 


■ | 


o 


H 


2 








o 






O 


o 


o 


rH 


£5 






















«-l 


<J 






















a) 


Q 






















OJ 


< 






o 


o 


ro 


O 


CN 


O 


u-i 


oo o 


n 


Q 




c 














iH 




j-i 


o* 


















CN 




OJ 
























X 




















00 


o 


•u 




3 




J 


1 

1 


1 
1 


1 
1 


1 

1 


1 
1 


vD 


m i 
• i 


c 




W 
















i— I 


CM 


o 

























*COH(NOlflOCOM^inCl 
r—i -cf CO vO *X 

iH Cn OJ ,-H 



r>- O <f rH O iH 

■Hoi i i <r ' cn ^-i- o i I 

• •III • I • • . I I 

rH O O rH O O 



-jOvOinoo-jajONcio 

rH r^ CI CN 



CN 


>* 


o 


c> 

o 


o 


o 



I I CN I I 



I I 



I I 



o o o vo 

O CI o O 

o" o' d o* 



I i i 
I I I 



o 
o 



OOrOrHOOOOCNu-lrHO 

vO CN r-. rH 

CN ^H 



I I O 



.O I 



■J" CI 

I O O I 

'o'd' 



u-i 
O 



O 
O 



O HOO CA O •* "1 
■H CN u~l CJ\ -J- 



cn rH r-» ci ui 

I I fl HO\ HN 



vO O 

■it 



O O O —i o 



OH-JHNOCOrMWvCNO 
lAHM-J H CNClOCNiH 



C^ClOOOrHOCN-^l^JHO 

inoomp-iciu-ioooiHO 
ooooooooooo 



o 
o 



d 



TO 
go 
o 






















































E 


rH 


















OJ 


CU 


















CJ 


4-1 










0) 








CO 


CO 










N 




• fH 




U 


o 










•H 




C rH 




« 


• 








o 




14 OJ 


• 


• 


X >^ 








> 




< X 


X 


l-l /-N 


-4 OJ 


• 






c • 




fa 


CO 


CO • 


ai • >h 


M 






CU c 




K 


O 


> X 


•H Jrf 03 


CO 






OS u 




OJ <~^ • 


•fa' CU 


X 


> rH e 


> 






OJ 




• hH 


>-. u u 


• o 


T3 • • CU 








CO 


•H 




• B co 


OJ ^n\Xu m 


T3 OS 


^-v CO J D. X 


• 






H 


X 




MS S 


S • X 


rH -W 


• as co 


J 


• 




re 






'- — o 


M - <~* 


^H 


^ CO CO ^~> 




E 




X 


CO 






S N-' X 


• CO rJ 


■H CO 




CO CO 


■H t. 


E 




CO 


• 


a 


1- 




• 




fa SQ ^ 


S OJ 




>- 3 


rH 


• CU 


3 




• kJ 


E 


OJ 


O 




4-1 


• CO 


E 


s^ 1-1 


• co 


-o 


(J 


O CJ 


a 


■J J«! rH 


iH 




J 


CO 


CJ 


H 




CO 


W -H 


3 





•H 


X « 


CO 


•H 


<y 


fa -H 


IH 


CO X 


rH 




CO 


J2 


J 


O 


<4-l 




u 


CO 


-3 


3 





X 


CO w 


05 


O 


5 


w lJ 


■H 


OJ 


m co 


5% 




re 


•H 


c 




u 


•H 




o 


co 


3 


4-1 


u 


3 


CJ 


o 


O 


CU 


CO 


X 


4J 


E 


v_^ > 


-3 




S: 


rH 


3 


cd 


u 


4J 


• 


'j- 


±j 


CU 


3 


-C 


U 


4-1 


fa CO 


£ 


C 


pq 


CO 


re 


OJ 


h 


a] 


a. 




o 


o 


X 


X 


•- ' 


iH 


H 




CO 


— i 


CO 


CO 


o 


c 


X 





3 




c 


-o 


CO 


01 


CO 


cu 





X 


•- 


IK 


H 


4-1 


E 


3 


pq 


CO 


•H 


re 


U 


H 


H 


c 


•H 


4J 


u 


E 


■H 


—i 


•rH 


c 


•H 


a 


3 


X 


CJ 


tH 




3 






rH 


T3 


M 


o 


OC 


uh 





CU 


3 


CO 


a. 


3 


h 


CO 


3 


•H 


X 


•H 


•H 


o 


u 


V4 


CO 


CO 


CO 




•H 


3 


•H 


CU 


CO 




•H 


re 


N 


CO 


u 


o 


h 


co 




cr 




CU 


Vj 




fa 


Bi 


>^ 


X 


CJ 


s^ 


CO 


> 


T3 


E 


c 




E 


4J 


ij 


■-H 


CJ 




a. 


CO 





DO 


01 


C 


iH 


oj 


E 




CO 


a 


CU 


CO 


X 


■- 


CO 


•H 


•H 


CU 


c 


3 


3 







■H 


CO 


4J 


3 


o 


X 


0) 


3 


CO 




H 


u 


CJ 


u 


1 


CJ 


•o 


cn 


o 


M 


3 


CO 


> 


l-l 


u 


CO 


3 


CO 


iH 


CO 


rH 


C- 




iH 


a 


CO 


CO 


<4-l 


4-1 


CO 


V 


■H 


CO 




H 


T3 


1 


H 


3 


4-1 


OJ 


•H 


CO 


•H 


H 


C 


>, 


o 


CO 


—i 


CO 


T3 


•H 


CU 


cu 


4-1 


CO 


> 




CO 


■n 


>^ 


3 


4-1 


OJ 




N 


CO 


3 


3 


>. 


■H 


X 


4-1 


H 


>^ 


E 


M 


-3 


i-i 


4-1 


DO 


X 


CO 


BO 


CU 


3 


X 


CU 


h 


l-l 


CO 


4J 


CU 




3 


X 


h 





a. 





— 


TO 


- 


u 






o 


a 




T4. 


4-1 


3 


o 


-a 


• 3 


. c 


■H 


c 


a 


OJ 


c_ 


CO 


u 


>, 


> 


C_ 


j; 


4-1 





co 


CO 


l-l 


CO 


CO 


X 


>^ 


OJ 


X 


o 


W O 


- c 


C 


co 


3 


CJ 


o 


3 


<u 


h 


OJ 


o 


CO 


4-1 


X 


3 


E 


X 


>i 


3 


a 


■a 


4J 


4-1 


u 


CO CJ 


CO OJ 


E 


CU 


CJ 


•H 


rH 


BO 


-a 


o 


CO 


H 


4-1 


CU 


a, 


CJ 


•H 


CJ 


iH 


CJ 


E 


o 


i^ 


>. 


CU 


> CO 


> 1- 


a 


X 


•H 


> 


r-j 


re 


»H 


u 


CO 




3 


3 


•H 


N 


■H 


o 


•H 


CD 


o, 


CO 


1^ 


^H 


rH 


CO 


►J 


H 


fa 


< 


< 


o 


cn 




C/0 


U 




u 


U 


fa 


<: 


a 


Pw 


fa 


fa 


< 


z 


fa 


CJ 




fa 




H 



11 





CO 






^N 


* 






c 




• /-N 




co 




M en 




M) 




cfl r~» 




01 




m en 




>H 


cfl 




v^ ' 




0) 


c 








C 




a 


• • 




a 


CD 




3 


a '-n 




3 


(-i 




4-1 


efl O 




c 


CO 




CO 


r4 <" 




■H 


H 




cu 


^x 









4-> 


CO 






CO 


n 


o 


•H 


• 




cd 


/~N 


CO 


n 


X 




"-5 


00 


U 


cfl 


Cfl 




v£> 


X 


■H 


o 




•* v^ 


•H 


rH 


Fn 




/~\ 


X 


•rH 






CN • 


CJ 


cfl 




<r c 


a 


c 




v^ n 


3 


CO 


Cfl 




CU 


X! 


a 


•H 




Vh -H 


4-1 


X 


4-1 




0) PS 


c 


o 


r4 




^ 


CO 


o 


3 




CO CO 




o 


o 




m c 




- CJ 




CO 




/-^ /-N 'O 




cfl 


Vj 




en -3- M 




•H 


cu 




en ^-i cu 




<H 


TJ 




^^ >-• > 




o 


s 




• 




■4-1 


o 




60 • cfl 




•H 


CO 




s a b 




4J 




O eo C 




CO 


CO 




hi-l ai 




3 


■H 




m m 




60 


CO 




CO 


cfl 




(3 


>•» 




^-s 4J 


H 




CO 


H 




rH CO 






CO 




CTl T3 


» 




• CJ 




w H 


^\ 




H -H 




O 


rH 




CO »-i 




• o 


r^ 




> H 




XI X 


N— r* 






CJ 3 






• #» 


4-1 CO 


• 




s ^ 


3 rH 




CO 00 


ps co 


3 




►J o> 


•H 


H 




CO 


CO 


T3 






CO 


^^ 


01 


u 


■rH 




X 




T3 


o 


■H 




3 


CO 


•rH 


u 


3 




H 


cu 


O 


•H 




4-1 


o 


rH 


•> > 




01 


(J 


>> 


'-N tH 




H 




X 


en O 




"•"N 


o 


m x 




CO 


• 


Cfl 


\s 




C 


X 


rH 


• efl 




Q) 


C 


CJ 


60 -H 




cO 


3 


B rH 




CJ 


X 


cfl 


O CO 




CO 


H 


X 


u c 




)-l 


>^ • 


a 


m -h 




P 


^* 


►. 


a 




a 


00 


rH 


"-N U 




3 


^o 


eo 


• cu 




•• CJ 


H 


CJ 


rJ H 




■H 


^^ 


o 


^x 




*-\ P- 




•» 




G O 


^ 


cfl 


^^ 




w iH 


0) 


•• CJ 


LO 




X 


a 


•H 


rH 




iH 4-1 


cu 


O 4-1 


^^ 




O <D 


n 


O cfl 






• CO 


M 


• -H 


• 




O 




O CO 


4H 




c 


^^ 


cfl 


• 




4-1 O 


• 


4-1 rj 




O rH 


tH 


O cfl 






r» 


CO 


c 


cfl 




(3 X 


a 


B -H 


CU 




o o 


0) 


O U 


4-1 




•H M 


Hi 


•H X 


(3 




4-> 4J 


v^ 


4-1 3 


01 




CO CO 




Cfl rH 


60 


• 


4-> > 


, Cfl 


4-1 O 


U 


H 


co a 


•H 


CO U 


cfl 




0) ' u 


CU 


0) 


4-1 "CO 


4-1 "Cfl 

(3 "■» -H 


X 


•HvOH 


•H en 4J 


(0 


CO "H 


oi t-l 


H 


JC -^ X 


X ^ O 




4-1 CO 


4J 4-1 


O 


•H • 1 


•H • CU 


4-1 


13 iH cfl 


sec 




X U 


U Vj 


00 


cooo. 


co cu 3 


(3 


<U C 3 


CU "H O 


•H 


•H M CO 


•H PS H 


4J 


o 


o 


O 


0) 


cu 


O 


(X 


a 


fn 


V 


1 




V 


1 






UO ! J B J S3 J U | 



u o i ) e i s e ) u | 




a 




Q 




n 




c 




o 




«rl 




4J 




id 




o 




o 




rH 




4J 




X 




M 




•H 




■u 




■U 




03 




CO 




OJ 




■H 




a 




11 




a. 




n 




0) 




rH 




XI 




•H 




4J 




a. 




ai 




a 




en 




3 




(fl 




0) 




> 


• 


•H 


rn 


UH 


CO 




9i 


d 


rH 


o 


| 




o 


a 


CO 


o 


a-. 


•H 


tH 


1-1 




td 


•> 


4-1 


^ 


to 


H 


oj 


r-t 


14-1 


n) 


C 


3 


■rl 


C 




C 


13 


s 


•H 


•H 


O 


X 


O 







■O 


o 


01 




u 


c 


o 


cd 


1J 


•H 


•H 


T3 


a 


OJ 





S 





■=. 


rH 


■H 


iH 




o 


a) 


4-1 


60 < 


c 




fl 


(d 


£ 


rJ 


U 


XI 




id 


<u 


•O 


si 


t-H 


H 


<c 



60 
•H 



UO | JBJS8JU | 



uo 1 1 e j sa jui 



ATOLL RESEARCH BULLETIN 
No. 292 



SHORT ORIGINAL ARTICLES 



By 

Various Authors 



Issued By 
THE SMITHSONIAN INSTITUTION 
Washington, D« C-, U-S-A- 
May 1985 



EDITORS' NOTE 



In line with our policy of not issuing short 
articles as separate numbers with their own title 
pages, the following articles are offered as parts 
of a single number. 



Contents 



Page 

Non-selective fishing methods of Futuna (Horn 

Archipelago, West Polynesia), by Rene Galzin 1 

Croissance et production de Chama io stoma dans le lagon 
de Takapoto, Tuamotu, Polynesie francaise, 
by Georges Richard 11 

First records of Wood Sandpiper, Ruff, and Eurasian Tree 

Sparrow from the Marshall Islands, by Manfred Temme ... 23 

Classification of emergent reef surfaces, 

by F . R. Fosberg 29 

Botanical visits to Krakatau in 1958 and 1963, 

by F. R. Fosberg 39 

Checklist of the herpetofauna of the Mascarene Islands, 

by D. D. Tirvengadum and R. Bour 49 

Marine and terrestrial flora and fauna notes on Sombrero 

Island in the Caribbean, by Nancy B. Ogden, William B. 
Gladfelter, John C. Ogden and Elizabeth H. Gladfelter .. 61 

Vegetation and flora of the Lowendal Islands, Western 

Australia , by Ralf Buckley 75 

Notes on a brief visit to Seringapatam Atoll, North West 

shelf, Australia, by B. R. Wilson 83 

Sea snakes collected at Chesterfield Reefs, Coral Sea, 

by Sherman A. Minton and William W. Dunson 101 

The underwater morphology of Palmerston and Suwarrow 

Atolls , by J . Irwin 109 



NON-SELECTIVE FISHING METHODS OF FUTUNA 
(HORN ARCHIPELAGO, WEST POLYNESIA) 

by Rene Galzin (-J) 



ABSTRACT 

Futuna, a high volcanic Pacific island without a lagoon, is 
surrounded by an "apron reef" which emerges at low tide. During 
spring tides, this reef flat is subject to heavy exploitation from the 
island's people and domestic animals (i.e. pigs) through fishing, 
collecting shells, crustaceans, and echinoderms, and turning over 
stones and corals. 

We describe two non-selective fishing methods used by the 
island's women: application of futu (a toxic substance obtained from 
the seed of Barringtonia asiatica ), and construction of small rock 
piles to attract juvenile fish and to be dismantled after a day or two 
to collect the fish hiding inside. 

At three stations on the sites where these fishing methods are 
employed, we collected, through poisoning experiments, approximately 
40 species of fishes belonging to 20 families. For each species we 
give the number, average length and weight of the catch. In the 
species taken, the length ranges from 1.5 to 17.2 cm and the weight 
varies from 0.1 to 90.4 g. 

We feel that these two non-selective fishing methods may endanger 
the balance of tine ichthyological fauna of this island, as almost 58% 
of collected species were juveniles, e.g. Sargocentron rub rum, 
Epinephelus merra, Lut.ianus monostigmus, Halichoeres margaritaceus, 
Acanthurus triostegus, Ctenochaetus striatus, Naso unicornis . 

INTRODUCTION 

A study of the coral reefs and their potential resources was 
undertaken on Futuna (Horn Archipelago, 2000 km north-east of New 
Caledonia) in November 1980. 



(1) Laboratoire de Biologie marine et de Malacologie - Ecole Pratique 
des Hautes Etudes - 55 rue de Buffon - 75005 PARIS and Centre de 
l'Environnement de Moorea - Museum National d'Histoire Naturelle et 
Ecole Pratique des Hautes Etudes en Polynesie francaise - B.P. 12 - 
MOOREA, POLYNESIE FRANCAISE. 

Atoll Res. Bull. No, 292: 1-10, 1985. 



Futuna is a high volcanic island, 18 km long and 6 km wide, 
situated at 1i»°15' latitude south and 178°10' longitude west. There 
is no lagoon, but the ocean shore includes a fringing apron-reef in 
places almost exclusively made up of eroded calcareous pavement reef- 
rock. This reef flat, of narrow and variable width (maximum width 
500 m), is almost totally emerged at low spring tide, and is covered 
by less than 80 cm of water at high tide. The geomorphology, biology 
and socio-economy of the Futuna marine ecosystems have been discussed 
elsewhere by RICHARD et al. (1982). The geomorphological observations 
made on the fringing reefs at Poi and Toloke (Figure 1) are as 
follows. The inner Poi reef is a reef-rock pavement partly covered by 
sedimentary accumulations. Next on the seaward side is a reef flat 
with widely spaced transverse ridges and furrows and, further seaward, 
a slightly raised pitted area. Toward the ocean, there is an inner 
biogenic ridge with a fragmentary covering of Melobesia and beyond 
that, the spur and groove area. The Toloke reef comprises a 
reticulated reef flat with furrows edged with crown-shaped corals. 
There are megablocks in the innermost area, and, seaward, a raised 
central zone with several branching Madreporia and Melobesia , sloping 
at its outer edge to the furrowed area. 

The Futuna population (about 3,000 inhabitants) does not seem to 
live in close association with the sea. According to FUSIMALOHI and 
GRANDPERRIN (1980), this situation arose from cultural and religious 
prohibitions and taboos regarding the marine environment and dates 
back more than a century. GAILLOT (1961) thinks that the decline of 
fishing both on the reef and along the coast of Futuna today results 
from the regular importation of canned fish and meat. The local 
Economic Service (Services Territoriaux de l'Economie Rurale) believes 
that the appeal of the sea disappeared amongst Futunans due to 1/ the 
spread of a colonial -style prejudice toward the canoes previously used 
on the open sea; 2/ difficulties in the construction and maintenance 
of the very heavy, traditional canoes; and 3/ religious and tribal 
interdicts. For the year 1979, the Economic Service estimates the 
catch of fish at 32 metric tons (30 tons of ocean fish and 2 tons of 
reef fish). 

In this paper we describe the fishing methods used by the 
Futunans, emphasizing in particular two traditional methods which 
appear to endanger fish populations. These latter are stupefying fish 
with futu and trapping them in piles of stones. This is not the first 
time that these two methods have been described in literature on 
fishing in the Pacific islands. We can quote among others the works 
of STOKES (1921), BURROWS (1936), LEONARD (1938), KRUMHOLZ (1958), 
LEGAND (1950), GAILLAND (1961), RANDALL (1963), BAGNIS et al. (1973), 
GALZIN (1977), SALVAT et al. (1978). The biometric analyses of the 
fish caught on the reef were made from samples obtained by poisoning. 

FISHING METHODS USED ON FUTUNA 






Fishing on Futuna seems to have developed in three main stages. 
Originally, before the arrival of European explorers (the island was 
discovered in 1616 by the Dutchmen LEMAIRE and SCHOUTEN) and until the 
end of the 19th century, the Futunans, like other Pacific Maohi 
peoples, were probably expert in the use of reef and ocean resources. 
However, as pointed out by DOUMENGE (1966), one must be wary of 
attributing too readily to these Polynesians a natural vocation for 
the exploitation of the sea. The second stage spans the period from 
the end of the 19th to well past half of the 20th century. During 
this period, Futunans became almost exclusively farmers, this mainly 
due to the strong influence of Catholicism (Fusinalohi and 
Grandperrin, 1980). Finally, since 1970, the Economic Service has 
been trying a program of subsidies and boat construction to "reteach 
the sea" to the Futunans. 

Methods presently used on Futuna 

All methods, apart from troll line fishing, are used from the 
reef or on its immediate outer slope. 

Troll line fishing is mainly practised around the north cape of 
Futuna and the south cape of nearby Alofi island. Barracuda comprise 
more than 70% of the catch. Jacks, frigate mackerel, tuna, swordfish 
and dolphins are also fished. 

Dropline fishing on the outer slope has always been practised by 
Futunans from their small kumete (small Futunan canoes adapted 
originally from food bowls and without an outrigger). New deep-sea 
fishing techniques for catching snappers (vivanos) are currently being 
developed by the South Pacific Commission and 0RST0M (F0URMAN0IR, 
1980) with a view to teaching the local fishermen. At present, there 
are six boats equipped for such fishing in Futuna. 

Gill nets have always been used. They are now manufactured from 
synthetic material (nylon) and imported from Noumea. Collective 
fishing with long surrounding nets seems to have been more or less 
abandoned and replaced by permanently-anchored gill nets. For 
example, at Sigave Bay (Figure 1), the number of gill nets has 
increased in two years from about 10 to more than 50. Nets with a 
3 cm mesh can stay anchored for three to four months with catches 
comprising mainly kingfish Selar crumenophthalmus (locally known as 
atule ), jacks and mullet. 

Line fishing takes place by day and night on the reef slopes. 
This method is practised by both men and women, with the catch 
consisting mainly of jacks, squirrel fishes, and snappers. 

The first underwater harpoon was imported from Noumea in 1961. 
There are now nearly 100 in use by both men and women with the catch 
including mainly parrotfishes, jacks, surgeonfishes, rabbitfishes and 



the snapper Macolor niger . 

Fishing by torchlight is practised by women on the reef at low 
tide on nights without moonlight. Crustacea and shellfish are also 
caught by this method. 

Fishing with barbed spears is done by the men either on foot from 
the reef or by diving, but this is now dying out. 

Fishing with futu and fishing with piles of stones are two 
methods used by the women, and will be described in detail below. 

Fishing for f lying fish and fishing with the use of snares are 
two methods described by GREZEL (1878), but they seem no longer to be 
practised. 

Fishing with explosives ; we were unable to ascertain whether 
this method, frequently used on Wallis, and one of the reasons for 
the scarcity of fish in the lagoon of this neighboring island, is also 
practiced on Futuna. 

Fishing with futu 

This type of fishing observed at Poi, tends to prevent fish from 
escaping by stupefying them with a vegetable poison. Two or three 
times a week, at low tide, an area of the inner reef flat, covered by 
only a few centimeters of water, is surrounded with branches and dry 
leaves of coconut trees. A few days in advance, about 20 Barringtonia 
asiatica fruits are collected and their big, round seeds grated. The 
powder obtained by grating the Barringtonia seeds is put in a basket 
which is then submerged and shaken within the enclosed area. About 
half an hour later, the fish begin to flounder and break surface - 
they are then collected with a kind of skimming net called kukutsi . 

Trapping with piles of stones 

This method was observed at Toloke in an area of reticulated reef 
with residual pools and puddles. These depressions (3 m wide by 150 m 
long and 0.40 m deep) have a sandy bottom over the reef pavement. The 
women of the village build an artificial shelter from a pile of 
stones, about 1 sq. m in size in one part of a basin, and this is left 
for a few days without being touched. On fishing day, a basket woven 
from coconut leaves is placed between the pile and the edge of the 
basin, and then the stones are removed one by one from the side away 
from the basket. Two or three of them are put in the basket to 
constitute another precarious refuge and the rest are put in the other 
part of the depression. The fish, which had found shelter within the 
pile, gradually see their refuge getting progressively smaller. All 
or nearly all of them seek their last shelter under the two or three 
stones in the basket. These stones are then removed and the fishing 



process is complete. 

The two methods just described above have two characteristics in 
common; the catch is poor for the amount of work involved, and only 
small specimens can be caught (the maximum size of those taken during 
our observation was 17 cm). A list of all species thus collected is 
given in Table 1. 

POISONING WITH ROTENONE POWDER 

Not wishing to deprive the Futuna women of their catch, yet 
wanting to check the biometrical parameters of the fishes thus caught, 
we collected fishes with rotenone (GALZIN, 1979) in the same areas 
several days later. The results of these catches are summarized in 
Table 2. At Poi, near the beach, 102 fishes were caught, with a total 
weight of 153 g, i.e. 1.5 g per individual. The largest fish caught 
was less than 10 cm long. On the same reef, but along the biogenic 
ridge, we caught 94 fishes with a total weight of 237 g (average 
weight 2.52 g). Here again, the biggest fish (Halichoeres 
margaritaceus ) was very small (10 cm long and weighing 19.5 g ). At 
Toloke, the fishes caught were somewhat bigger. In a depression of 13 
sq. m we caught 139 fishes with a total weight of 1031.8 g - an 
average weight of 7.2 g each. The largest was a serranid - 
Epinephelus merra T weighing 90.4 g and 17.2 cm long. 

These figures give an idea of the size of fish caught at low tide 
by the women of Futuna who gather all they can find on the reef. 
Among the fishes caught (Table 1) at least 26 species (58%) were at a 
juvenile stage: included were, e.g. Sargocentron rubrum, Epinephelus 
melanostigma, £, merra, Lutianus monostigmus, Halichoeres 
margaritaceus, Acanthurus trios tegus, Ctenochoetus striatus and Naso 
unicornis . 

CONCLUSIONS 

At low tides, the reef flat is a "rendez-vous" for both the human 
and animal populations of the island. The inhabitants collect all 
that is edible, while pigs rummage and turn over every stone, causing 
the destruction of sciaphile flora and fauna. However, in the absence 
of quantitative information on harvest rates, it is not possible to 
confirm the occurrence of over-harvesting. 

With the two traditional fishing methods described above (futu 
and stones), the fish caught are almost exclusively of very small size 
(as noted by BURROWS, 1936). These methods thus tend not only to 
cause the disappearance of sedentary fish populations of the reef 
flat, but also to contribute to reduce fish stocks living on the outer 
slope: juveniles of most of the fish species living as adults on the 
outer slopes have to find shelter in the calmer and more trophic 
environment of the apron-reef where they are harvested 



indiscriminately. More information is needed on the behavior of all 
fish species as the tides go down. 

Considering the current Futunan population size, with the 
survival of ancient techniques of fishing as well as the introduction 
of new fishing methods, we observe a considerable non-selective 
exploitation of the apron-reef fishes on Futuna. An information 
campaign concerning the problems of maintaining natural stocks of fish 
should be launched to complete the excellent initiative already 
undertaken to promote and develop offshore fishing. 

ACKNOWLEDGEMENTS 

We thank A. Mauge for assistance in identifying taxa and J. 
Newhouse, J. Randall, Patrick V. Kirch and M.-H. Sachet who provided 
critical comments and suggestions on earlier drafts of the manuscript. 

This study was made possible by contract N° 02565 between the 
French Territory of Wallis and Futuna and the Association Naturalia 
and Biologia. 

REFERENCES 

BAGNIS, R., BOULIGUEUX, G., DROLLET, J. & PETARD, P., 1973 - Poisons 
vegetaux de p£che des anciens polynesiens. SPC, 
Medicinal Plants, W.P., 29 : 16 p 

BURROWS, E.G., 1936 - Ethnology of Futuna. Bernice P. Bishop Museum 
Bulletin, 145 : 176 p. 

DOUMENGE, F., 1966 - L'homme dans le Pacifique sud. Publication de 
la Societe' des Oce'anistes, 19 : 633 p. 

FUSIMALOHI, T. & GRANDPERRIN, R., 1 980 - Rapport sur le projet de 
developpement de la peche profonde a Wallis et Futuna. 
South Pacific Commission, 861 : 25 p. 

FOURMANOIR, P., 1980 - Mission a Wallis et a Futuna pour la peche 
profonde des vivanos rouges (Etelis) a la palangre. 
Rapport 0RST0M : 7 p. 

GAILLOT, M., 1961 - Un type de peche dans le Pacifique: la peche a 
Futuna. Les cahiers d'Outre-Mer, 55 : 317-322. 

GALZIN, R., 1977 - Richesse et productivite' des ecosystemes 

lagunaires et re'cifaux. Application a l'etude dynamique 
d'une population de Pomacentrus nigricans du lagon de 
Moorea (Polyne'sie frangaise). These de specialite', 
U.S.T.L. Montpellier, 109 p. 27 fig., 30 tabl.. 



GALZIN, R., 1979 - La faune ichtyologique d'un recif corallien de 

Moorea. Polynesie francaise: Echantillonnage et premiers 
resultats. Terre Vie, Rev. Ecol., Vol.33 : 623-643 

GREZEL, R. Pere, 1878 - Dictionnaire Futunien-Francais. Paris. 

KRUMHOLZ, LA., 1948 - The use of rotenone in fisheries research. 
Jour. Wildlife Mangmt., 12: 305-317 

LEGAND, M., 1950 - Contribution a l'^tude des methodes de peche dans 
les territoires francais du Pacifique Sud. Jour. Soc. 
Oceanistes, 6 (6) : 141-184 

LEONARD, J.W. , 1938 - Notes on the use of derris as a fish poison. 
Trans, of the American Fish. Society, 68 : 269-279 

RANDALL, J.E. , 1963 - Methods of collecting small fishes. 
Underwater Naturalist, 1 (2) : 6-11 

RICHARD, G., GALZIN, R., SALVAT B., BAGNIS R., BENNETT J., DENIZOT, 
M., & RICARD, M., 1982 - Geomorphology, ecology and 
socio-economy of the Futuna marine ecosystem (Horn 
Archipelago, Polynesia). IV Intern. Coral Reef Symp., 
Manila, 1 : 269-274 

SALVAT B., RICARD, M. , RICHARD, G., GALZIN, R., & TOFFART, J.L., 1 978 - 
A summary review of the ecology and reef-lagoon economy 
of Lau. UNESCO/UNFPA Fiji Island Reports, 4 (Canberra, 
ANU for UNESCO): 129-145 

STOKES, J.F.G., 1921 - Fish poisoning in the Hawaiian Islands. Occ. 
Paper Bishop Museum, 7 : 219-233 




40 (0 

•rH 

V s 



•HClO 
O -H jj 

x! dj 

CI -H 

q aj <u 

■3 OH 

x: ?-, 3 

to +J 
O M 

x: p -3 
■u P x: 

O 3 Kh 

O <V-,| 
•H 0) 

p£ > 
(X) 



B£. 



(0 >> 

<Vh 0) ••> 

o S o) 

0) 
D.-H ^H 
Cfl O -H 
S O. D. 










ROTENONE 










EXPERIMENT 


3 
3 


u 

o 


a> 


a 








i_ « 

C i. 


*" JC 

is rt 
a o> 


UJ 

o 

—1 

o 


i 

St 

c 


' 0) 
O) ~ 

c a 






5 

a. 


5 

O- 


1- 


IE 
id 

iZ 


(A 


MURAENIDAE 


Echidna nebulosa (Ahl,1789) 
Lycodontis sp. 






+ 




+ 


CONGRIDAE 


Conger cinereus Ruppell,1828 


+ 










HOLOCENTRIDAE 


Sargocentron rubrum (Forskall,1775) 

" en 








+ 


+ 


SYNGNATHIDAE 


sp. 
Choeroichthys sculptus (Giinther,1870) 


+ 










SCORPAENIDAE 


Sebastapistes corallicola Jenkins, 1902 






+ 








Scorpaenodes guamensis (Quoy et Gaimard,1824) 


+ 


+ 






+ 


SERRANIDAE 


Epinephelus melanostigma Schultz,1953 
" merra (Bloch,1793) 




+ 


+ 


+ 


+ 


GRAMMISTIDAE 


Grammistes sexlineatus (Thunberg,1792) 


+ 


+ 


+ 






PLESIOPIDAE 


Plesiops caeruleolineatus (Ruppell,1835) 






+ 




+ 


APOGONIDAE 


Apogon aureus Lacepede,1803 










+ 




" cyanosoma (Bleeker,18S3) 


+ 




+ 




+ 




" nubilus (Carman, 1903) 






+ 




+ 


LUTJANIDAE 


Lutjanus monostigmus (Cuvier,1828) 








+ 




MULLIDAE 


Parupeneus atrocingulatus (Kner,1870) 






+ 






CHAETODONTIDAE 


Chaetodon citrinellus Cuvier,1831 






+ 








11 lunula (Lacepede,1802) 


+ 


+ 




+ 






Pomacanthus imperator (Bloch,1787) 


+ 










POMACENTRIDAE 


Abudefduf sordidus (Forskall,1775) 
Stegastes nigricans (Lacepede,1803) 
Chrysiptera cyanea (Quoy et Gaimard,1825) 




+ 


+ 
+ 




+ 




11 glauca (Cuvier,1830) 


+ 


+ 


+ 


+ 






" leucopoma (Lesson, 1830) 


+ 




+ 








Pomacentridae sp.1 (juv.) 




+ 










" sp.2 (juv.) 






+ 






LABRIDAE 


Halichoeres hortulanus (Lacepede,1802) 






+ 








margaritaceus (Valenciennes, 18 39) 


+ 




+ 


+ 






" marginatus (Riippell,1835) 






+ 








Stethojulis trilineata (Bloch et Schneider, 1801) 


-^. 




+ 


+ 






sp. 
Thalassoma hardwickei (Bennett, 1830) 






+ 








" umbrostygma (Ruppell,1835) 


+ 




+ 






BLENNIIDAE 


Cirripectes variolosus (Valenciennes, 1836) 
Istiblennius cyanostigma (Bleeker,1849) 
" edentulus (Schneider, 1801) 
' ' periophthalmus (Valenciennes ,1836) 


+ 


+ 


+ 
+ 


+ 
+ 




TRIPTERYGIIDAE 


Tripterygiidae sp.(juv.) 


+ 










GOBIIDAE 


Bathigobius fuscus Riippell,1828 


+ 






+ 




ACANTHURIDAE 


Acanthurus nigrofuscus (Forskall,1775) 
" lineatus Linne,1758 






+ 


+ 






" triostegus (Linne,1758) 


+ 


+ 


+ 


+ 


+ 




Ctenochaetus striatus (Quoy et Gaimard,1824) 










+ 




Naso unicornis (Forskall,1775) 










+ 


Number of species 


16 


9 

* 


25 


12 


12 



Table 1. List of fishes collected on the reefs of FUTUNA island. 







4-1 
O 

S-i CD 


WEIGHT in g 


SIZE in cm 




01 
b0-l-> 


§ +- 1 


E +-• 


bO 


| 


| 






CD -H 


oj JZ 


6 x: 


% xi 


cfl 


£ 


c 






£> O 


!-i ao 


•h bo 


•3 bO 


5h CD 


•H CD 


• H CD 






6 <u 


CD ■!-{ 


c; -H 


3 CD 


CD N 


C w 








3 p 


> CD 


• H CD 


> -H 


• H -H 






z <n 


rt 3 


e 2 


e 3 


n5 m 


£ <f> 


£ <s> 




Conger cinereus 
Choeroichthys sculptus 


1 


0.8 
0.2 


0.6 


1.0 


8.3 
4.1 


7.0 


9.7 


POI 


Scorpaenodes guamensis 


1 


0.6 






3.0 








Grammistes sexlineatus 


2 


0.4 


0.2 


0.7 


2.7 


2.1 


3.2 




Apogon cyanosoma 


3 


2.5 


1.6 


3.3 


5.5 


4.8 


6.1 


NEAR 


Chaetodon lunula 
Pomacanthus imperator 


2 
1 


3.9 
1.3 


0.6 


7.3 


4.2 
3.2 


2.4 


6.0 




Chrysiptera glauca 


25 


3.8 


0.2 


12.1 


5.2 


2.1 


8.2 


THE 


leucopoma 


2 


0.7 


0.5 


0.9 


3.1 


2.7 


3.5 




Halichoeres margaritaceus 


6 


6.3 


0.4 


19.5 


6.4 


3.0 


10.0 




Stethojulis sp. 


1 


0.7 






3.5 






RIDGE 


Thalassemia umbrostygma 


3 


0.2 


0.1 


0.3 


2.4 


2.0 


3.0 




Istiblennius periophthalmus 


4 


0.8 


0.5 


1.2 


4.1 


3.5 


4.7 




Tripterygiidae sp. 


2 


0.1 


0.1 


0.1 


1.9 


1.6 


2.1 




Bathygobius fuscus 


20 


1.7 


0.1 


6.9 


4.9 


2.3 


8.6 




Acanthurus triostegus 


19 


2.3 


0.6 


5.0 


4.3 


2.8 


5.7 




Scorpaenodes guamensis 


2 


2.4 


2.1 


2.7 


5.1 


4.8 


5.3 


POI 


Epinephelus melanostigma 


5 


2.7 


1.2 


6.3 


5.5 


4.3 


7.8 




Grammistes sexlineatus 


4 


4.9 


0.3 


17.5 


4.6 


2.2 


9.3 


NEAR 


Chaetodon lunula 


1 


0.2 






1.9 








Abudefdut sordidus 


2 


0.6 


0.5 


0.7 


2.8 


2.7 


2.9 


THE 


Chrysiptera glauca 
Pomacentridae sp. 


40 
1 


1.5 
0.1 


0.1 


11.3 


3.7 
1.9 


1.9 


8.1 


BEACH 


Istiblennius edentulus 


1 


2.0 






5.8 








Acanthurus triostegus 


45 


1.1 


0.5 


6.5 


3.2 


2.6 


6.7 




Lycodontis sp. 


4 


6.9 


1.2 


16.5 


13.5 


8.4 


17.2 




Sargocentron sp. 


3 


3.5 


2.7 


5.0 


5.7 


5.1 


6.6 




Sebastapistes corallicola 


1 


5.8 






6.0 








Epinephellus merra 


4 


66.0 


34.6 


90.4 


15.6 


12.8 


17.2 




Grammistes sexlineatus 


1 


0.3 






2.6 








Plesiops caeruleolineatus 


1 


1.6 






4.8 








Apogon cyanosoma 


9 


14.3 


3.1 


11.8 


7.7 


5.6 


9.0 




" nubilus 


1 


7.7 






6.6 








Parupeneus atrocingulatus 


1 


3.5 






6.7 








Chaetodon citrinellus 


4 


9.1 


4.4 


13.6 


7.0 


5.7 


8.0 




Stegastes nigricans 


38 


4.7 


0.2 


18.5 


5.5 


2.0 


8.7 


TOLOKE 


Chrysiptera cyanea 


3 


0.1 


0.1 


0.2 


2.1 


2.0 


2.5 




glauca 


3 


3.8 


2.3 


6.8 


5.6 


4.8 


7.1 




leucopoma 


1 


0.1 






2.0 








Pomacentridae sp. 


1 


0.1 






1.9 








Halichoeres hortulanus 


1 


5.1 






7.3 








margaritaceus 


3 


1.8 


0.5 


2.6 


4.9 


3.2 


6.2 




" marginatus 


6 


1.2 


0.1 


4.1 


3.5 


1.7 


6.4 




Stethojulis trilineata 


7 


5.3 


1.0 


12.4 


6.8 


4.2 


9.2 




Thalassoma hardwickei 


3 


8.2 


6.6 


9.1 


8.2 


7.5 


8.7 




umbrostygma 


8 


16.8 


4.7 


25.8 


10.3 


6.H 


12.0 




Cirripectes variolosus 


1 


0.2 






2.4 








Istiblennius periophthalmus 


12 


6.5 


0.7 


12.8 


8.5 


4.0 


12.5 




Acanthurus lineatus 


3 


21.2 


15.0 


27.0 


10.2 


9.8 


10.5 




" triostegus 


20 


2.9 


0.7 


8.1 


4.7 


3.2 


7.0 



Table 2. Rotenone experiments in 3 areas of the FUTUNA reef. For each 
station we give the number of fishes caught, the average weight 
and size of fishes, and the minimum and maximum weight and size 
of fishes. 



CROISSANCE ET PRODUCTION DE CHAMA IOSTOMA DANS LE LAGON 
DE TAKAPOTO.TUAMOTU.POLYNESIE FRANCAISE 

by Georges Richard 



ABSTRACT 

Ckama ioAtoma Conrad is a species characteristic of many 
mollusc communities in lagoons of high islands and especially of 
closed atolls, in French Polynesia. 

Tagging of 40 Ckama at 4 sites led to an estimate of the 
growth of this species, and calculations from counts along 8 transects 
gave for the Ckama population of the entire Takapoto lagoon a total 
of about 11 million individuals (between 1 and 2 individuals/square 
meter in colonized areas). Analyses of population structure and 
a study of total weight show that the estimated 11 million indivi- 
duals represent a standing crop of 2,000 metric tons of total fresh 
weight including about 80 t of soft parts (corresponding to a mean 
soft biomass of 60 kg/hectare/year for densely populated areas, 
or 7.8 kg/ha/year for the whole lagoon bottom). The results of 
measurements made during two 8 months periods 6 months apart give 
a basis for calculation of a theoretical potential production of 
16 tons of soft biomass per year for the entire lagoon, or 12.5 
kg/ha/year for the densely colonized areas. 

Ckama LoAtoma is characterized by very slow growth rate, 
large standing crop, rather low productivity, and low P/B ratio. 
It belongs to a group of species of mediocre productivity, such 
as T/iidaana mwcuna and A/tca veruOvico/ia, already studied in French 
Polynesia. 



INTRODUCTION 

Dans les etudes de croissance et de production, les especes 
tropicales suscitent, depuis quelques annees, un interet qui va 
grandissant (RICHARD, 1982) . C'est le cas en Polynesie francaise, 
oG de nombreux travaux (RICHARD, 1977, 1978, 1981 ,1982a et b, 1983a 
et b, RICHARD et SALVAT,1982) analysent la distribution quantitative, 
la croissance et la production ou la productivity (potentiel de 
production) des especes les plus representatives de chaque grand 
type de formation recifale ou lagunaire. 

Laboratoire de Biologie marine et de Malacologie, Ecole Pratique des 

Hautes Etudes, 55 Rue de Buff on, 75005 Paris 
Antenne du Museum National d'Histoire Naturelle et EPHE, B.P. 12, 

Moorea, Polynesie francaise 
Atoll Res. Bull. No. 292: 11-22, 1985 



12 

Le present travail concerne le Bivalve Chamidae Chama LoAtoma 
Conrad, 1837, dans le lagon de 1 'atoll ferme de Takapoto, une des 
iles du Roi Georges, archipel des Tuamotu, Polynesie francaise. 



NORD 




OCEAN 



7LAVA71K.A 



FIGURE 1: Carte de 1" atoll de TAKAPOTO, montrant la position des 
prospections realisees sur Chama ^oAtoma, le long de 

la bordure lagunaire et sur les pates centraux. 



Les Chamidae, qui appartiennent a l'ordre des Hippuritoida, 
ont la particularity d'etre fixes au substrat par cimentation de 
l'une de leurs valves. Ce sont des coquilles epaisses, tres encrou- 
tees, inequivalves, solidement fixees, ce qui pose un delicat pro- 
bleme dans l'estimation de la taille et du poids. En Polynesie 
franchise, la famille des Chamidae est representee par 4 especes: 

Chama biaAALca Reeve, 1846 (Societe, Marquises) 

Chama paci.jU.ca Broderip, 1834 (Societe, Tuamotu, Gambier) 

Chama oApoyieAla Lamarck, 181 9 (Societe, Tuamotu, Gambier) 

Chama LoAtoma Conrad, 1837 (tous les archipels) 

Davantage par sa Constance que par sa reelle abondance, Chama 
lo/itoma est un element caracteristique des peuplements malacologiques 
des lagons d'lles hautes et, surtout, des lagons d'atolls fermes. 



13 

On la distingue aisement des autres especes de Ckama, par les taches 
violettes qui colorent l'interieur de ses valves. Nous analyserons 
successivement la croissance de cette espece, puis son bilan quantita- 
tif a l'echelle de tout le lagon de Takapoto (nombre d'individus 
et biomasse) et enfin sa production. Nous ferons suivre notre etude 
d'une breve comparaison avec les travaux precedemment realises 
sur d'autres especes polynesiennes. 



CROISSANCE 

A quatre stations, situees sur la bordure lagunaire (VA1RUA, 
TAKA1) ou sur les pates centraux (TARARO, OHAGUNA) du lagon de 
1' atoll ferme de Takapoto (figure 1), une quarantaine de Ckamaont 
LoAtoma ont ete mesures in situ (diametre de la valve libre), entre 
avril et decembre 1977, d'une part, et entre juin 1978 et fevrier 
1979, d'autre part (soit deux intervalles de 8 mois). Durant ces 
intervalles de temps, on releve des accroissements en taille variant 
de 1 a 6 mm, pour des Ckama loAtoma mesurant entre 50 et 53 mm 
au depart de l'experience. L'ensemble des resultats relatifs aux 
deux intervalles de temps nous permet de donner une expression 
de la croissance des Ckama LoAtoma (figure 2) qui obeit aux parame- 
tres suivants (equation de von BERTALANFFY,1938) : 

L = 86,9 (1 - e"°' 11t ) 

ou L = taille de la coquille au temps t. 

86,9 = Loo = taille maximum de la coquille, atteinte quand 
le taux de croissance est nul. 

t = age de 1' animal. (En fait, t = tx - to, to etant le temps 
auquel l'animal aurait eu une coquille de taille 
nulle; cette precision n'a pas de sens dans la presen- 
te etude et c'est pourquoi nous n'en tenons pas 
compte) . 

C'est ainsi qu'un Ckama LoAtoma dont le diametre de la valve 
libre mesure 13 mm a approximativement 1 an. Ceci traduit une vitesse 
de croissance tres lente (27,8% de Loo en 2 ans), mais toutefois 
moins lente que celle des A/ica verv&iLcoAa (17%) dans le meme lagon. 

11 nous a semble irrealiste de nous referer a la plus grande 
dimension de la coquille (valve inferieure fixee), pour les mesures 
in situ, et les donnees qui precedent concernent le diametre de 
la valve libre. Toutefois, il existe une relation lineaire entre 
le diametre de la valve libre (X) et la taille reelle (Y) des Ckama 

Y = 1,05 X + 10 (r = 0,9) 



14 



TAILLE 
(mm ) 



80 - 



60 . 



40 . 



20 




L«> = 86,9 



(r = 0,95) 



AGE (ans) 



FIGURE 2: Courbe de croissance de Ctujma -LOAtoma, calculee par 

la methode de von BERTALANFFY, atoll de Takapoto. 



BILAN QUANT1TAT1F NUMERIQUE 

Le bilan quantitatif numerique de la bordure lagunaire a 
ete jjresse a partir des 4 transects de ORAPA, VILLAGE, VA1RUA et 
GNAKE. L'ensemble des resultats, regroupes dans le tableau A, permet 
d'estimer le peuplement en Chama. -LoAtoma de cette bordure a environ 
3,8 millions d'individus. 





ORAPA 


VILLACE 


VAIRUA 


CNAKE 


TOTAL 


Longueur du 'transect' (m) 
Surface du "transect" (m) 
% de colonisation 
densite maximale (ind./m ) 


90 
180 
19% 
5,6 


160 
360 
22% 
3,8 


112,5 
225 
33% 
9 


137,5 
275 
27% 
8,8 


520 
1040 
25% 
9 


Nombre de Chama 

sur le 'transect" (indiv.) 
dans la zone (milliers) 


125 
975 


128 
371 


230 
1633 


156 
8^2 


6)9 
3820 


Den site par m 

sur le 'transect* 

dans 1 'aire colonisee 


0,69 
3,57 


0,36 
1 ,60 


1,02 
3,07 


0, 57 
2,06 





TABLbAL) A: Bilan des prospections quantitatives numeriques realisees 
sur la bordure lagunaire de Takapoto. 



15 

De la meme maniere, a partir des 4 transects de TARARO, TEAVAT1- 
KA, OHAGUNA et CENTRAL, dont les resultats sont regroupes dans 
le tableau B, le peuplement des 414 pates du lagon est estime a 
6,2 millions de Chama Lo/stoma. 





CENTRAL 


TEAVATIKA 


TARARO 


OHAGUNA 


TOTAL 


Longueur du transect (is) 
Surface du transect (m ) 
% de colonisation , 
densite maximale (ind/m ) 


102,5 
205 
73% 
3,6 


67,5 
135 

59% 
8 


75 
150 
50% 
2,* 


90 
180 
44% 
8,8 


335 

670 

8,8 


Nombre d'individus 
sur le transect 
sur le pate (milliers) 
sur la zone (milliers) 


155 

24,94 
673,380 


166 
17, 59 
2480,19 


91 

10,715 
2839,435 

0,61 
1 ,21 


156 

22,042 
815,554 


568 
75,29 

6208 


Densite au m 

sur le transect 

sur l'aire colonisee 


0,76 
1,03 


1,23 
2,08 


0,87 

1,95 





TABLEAU B: Bilan des prospections quantitatives numeriques realisees 
sur les pates centraux du lagon de Takapoto. 



Au total, en tenant compte des tests d'abondance realises 
sur le fond du lagon (densites de peuplement avoisinnant 100 indivi- 
dus/hectare au pied des pates, soit un stock de 900.000 individus), 
on aboutit a une population totale de 10,9 millions de Chama lo^toma 
pour l 1 ensemble du lagon de Takapoto. 



BIOMASSE 



Une etude des poids totaux, des poids de coquilles et de 
la biomasse, sur 3 populations types de Chama -lo-otoma, fait apparai- 
tre que le rapport de la biomasse au poids total est de 0,04, d'une 
part, et que le rapport du poids de la coquille au poids total 
est de 0,83, d'autre part. Le complement de poids (0,13%) correspond 
a l'eau intervalvaire. Le tableau C donne, a titre d'exemple, les 
mesures concernant la station GNAKE (avril 1977). 

Sur un echantillon de 100 Chama ia/stoma, la taille (diametre 
de la valve libre) est une fonction lineaire du logarithme de la 
biomasse, selon 1' equation de regression suivante: 

Y = 35,96 X + 32,61 

L'ensemble des donnees (bilan quantitatif, demographie, bio- 
masses) permet d'estimer que les 10,9 millions de Chama LoAtoma 
representent 1964 tonnes en poids frais total, dont 1630 tonnes 
de coquilles et 80 tonnes de biomasse. Cette derniere valeur corres- 
pond a une biomasse moyenne de 60 kg/ha pour la bordure lagunaire et 



16 

les pates, ou seulement 7,8 kg/ha si l'on se refere a toute la 
surface des fonds lagunaires. 

Dans 1' ensemble du present travail, comme ce fut le cas dans 
les publications precedentes relatives a la Polynesie francaise, 
craitant de l'un ou 1 'autre des aspects de la production en matiere 
vivante, nous entendons par biomasse le poids frais des parties 
molles de 1 'animal . 



TAILLE 


TAILLE 


P0I0S 


POIDS 


BIOMASSE 


1 valve 11 bre 


1 ongueur 


TOTAL 


C0QUILLE 




mm 


mm 


9 


P 


g 


60 


70 


lft3,0 


117,5 


3,5 


6ft 


60 


150,7 


121,7 


ft,9 


ft2 


ft9 


62,2 


ft8,2 


2,3 


33 


ft5 


35,0 


28,9 


0,9 


40 


ft6 


39,7 


36,2 


1.6 


ft9 


55 


72,0 


52,8 


2,5 


ft2 


50 


ft6,8 


35,8 


2,0 


5ft 


6ft 


106,9 


83, ft 


3,7 


66 


78 


21 9,0 


183,9 


7,9 


39 


55 


72,9 


57,9 


1,8 


60 


77 


167,0 


159,5 


7,1 


59 


68 


112,5 


86, 7 


ft, 9 


63 


72 


166,2 


1 5ft, ft 


6,8 


63 


80 


229,9 


190,2 


9,2 


72 


8ft 


2ftl,8 


201,9 


7,3 


52 


65 


103, ft 


81,1 


ft, 8 


ft5 


58 


68,5 


71,6 


3,7 


69 


89 


288, 7 


23ft, 5 


12,9 


ft5 


60 


107,6 


9ft, 1 


ft.l 


6ft 


69 


1 58,1 


131,0 


ft, 7 



TABLEAU C: Etude des poids totaux, des poids de coquilles et des 
parties molles sur une population de Chama Lo^toma, station 
GNAKE, atoll de Takapoto, avril 1977. 



PRODUCTION 



Des mesures effectuees sur 200 Chama -co^toma, situes a trois 
niveaux bathymetriques de deux stations de la Dordure lagunaire 
et des pates centraux (surface, -7 m de profondeur, -20 m de profon- 
deur), nous permettent d'etablir la structure demographique theorique 
des 10,9 millions d'individus du lagon. 

A partir des donnees de croissance (figure 2) et de biomasse, 
nous avons calcule 1 'accroissement theorique en biomasse des Chama 
LoAtoma pour un intervalle de temps de 8 mois, pour chaque classe 
de taille separement. Au total, cet accroissement ponderal est 
egal a 10,8 tonnes pour 8 mois (tableau D) et, en supposant la 
croissance peu variable dans le temps, a 16,2 tonnes par an. 

Les quelques donnees rassemblees sur le recrutement et la 
mortalite nous font considerer que la production theorique potentiel- 



17 

le de Chama LoAtorna correspond a 1 'accroissement ponderal des stocks 
de cette espece dans le meme intervalle de temps. Ceci correspond 
done a une production de 12,5 kg (biomasse) par hectare et par 
an pour les zones fortement colonisees, ou encore a une production 
de 1,6 kg/ha/an si l'on prend en compte tout le lagon. 



TAILLE 


TAILLE 


BIOMASSE 


BIOMASSE 


A8 


NOMBRE 


iB 


(T) 


(T + l) 


m 


(T+l) 


lndi vidueJ 


DE Chama 


TOTAL 


mm 


mm 


9 


9 


9 


i ndl vi dus 


kg 


15-20 


24,44 


0,38 


0,59 


0,21 


54.500 


11 ,45 


20-25 


28,94 


0,52 


0, 79 


0,27 


54. 500 


14,72 


25-30 


33,44 


0,72 


1,05 


0,33 





' 


30-35 


37,94 


0,99 


1,41 


0,42 


436.000 


183,12 


35-W 


42,44 


1,37 


1,88 


0, 51 


272.500 


138,98 


40-45 


46,94 


1,88 


2,50 


0,62 


872.000 


540,64 


45-50 


51,44 


2,59 


3,34 


0,75 


654.000 


490,50 


50-55 


55,94 


3,57 


4,45 


0,88 


1.144. 500 


1007,16 


55-60 


60,44 


4,92 


5,94 


1,02 


1.362. 500 


1389,75 


60-65 


64,94 


6,78 


7,93 


1,15 


2. 398.000 


2757,70 


65-70 


69,44 


9,34 


10,57 


1,23 


1 .758.500 


2212,1,6 


70-75 


73,94 


12,86 


14,10 


1,24 


1.199.000 


1486,76 


75-80 


78,44 


17,71 


18,91 


1,20 


381.500 


457,80 


80-85 


82,94 


24,40 


25,10 


0,70 


163.500 


114,45 


85-90 


- 


33,61 


- 


- 


54.500 


- 


90-95 


- 


46, 29 


- 


- 


54. 500 


- 










TOTAL: 


10.900.000 


10805 kg. 



TABLEAU D: Donnees permettant d'etablir 1 'accroissement en biomasse 
des 10,9 millions de Chama -LoAtoma, en huit mois, dans 
le lagon de Takapoto (annees 1978 - 1979). 



C0MPARA1S0N AVEC D'AUTRES ESPECES 

Que l'on considere le nombre d'individus ou la biomasse de 
n'importe lequel des milieux represented dans l'ecosysteme recifal 
polynesien, sa richesse est toujours le fait d'un nombre tres reduit 
d'especes. Dans le cadre de recherches visant a etablir la productivi- 
ty des complexes recifaux de cette region, ce sont justement ces 
quelques especes qui ont fait l'objet d'etudes sur la croissance 
et la production. 



Dans le lagon de Takapoto, cadre de la presente etude, deux 
autres Bivalves avaient fait precedemment l'objet de recherches 
tdentiques: T^Ldacna maxima (RICHARD, 1977, 1982a - HEN0CQUE.1980) 

et M.ca veruUvLcoAa (RICHARD, 1978) . Ces deux especes affichent une 
croissance tres lente et la vitesse de croissance de Chama LoAtoma 



18 

est intermediaire (% de Loo a 2 ans = 28) entre celle de T/iidacna 
mcocuna (41%) et celle de Aica vewUiLco/ia (17%). Les trois especes 
ont une production plutot faible, decoulant d'une part d'une forte 
biomasse et d'une forte production totale, mais, d'autre part, 
d'une croissance tres lente et d'un mauvais rapport Production/Bio- 
masse. 

Un autre Bivalve, Caa.cLLum ptag.um, a ete etudie de la meme 
maniere dans le lagon de ANAA. Cecte espece presente une croissance 
tres rapide, mettant seulement trois ans pour atteindre 95% de 
l°° , et les 600 millions d'individus recenses dans le lagon produi- 
raient annuellemenc 2200 tonnes de parties molles (RICHARD, 1982, a, b) . 

Pour la classe des Gasteropodes, les etudes de croissance 
et de production one jusqu'ici porte sur quatre especes: 
NojtLta pLLcata, sur un recif d'ilot d'une ile haute volcanique 
(Moorea-Societe) - 7 zcJicuiLua g/iandinatu^ , sur un recif exterieur 
d' atoll (Hao-Tuamotu) - OiostcuLca obvelata et M-LOia nvu&ia dans un 
complexe recifal d'ile haute (Moorea). Toutes ces especes ont des 
vitesses de croissance superieures a celle de Chama iojtoma; malgre 

cela, leur production est toujours inferieure, phenomene qui tient 
tantot a leur faible biomasse, tantot a un rapport P/B particuliere- 
menc bas. 



ESPECFS 


% de L» a 2 AUS 


Cirac tcr i s t iques de 
1 'e spece 


M 1 1. i r U 


L~ 


I Caldium liaqum 

? 1 c ci a * hi s Qiand inatu* 

1 ! . 

k £ 104 

■> ",/-,. ■-. 

('> T - 

7 Chama i o 

8 Atca rcntiicOAQ 




Bivalve endoge filtreijr 

' bopooe epige herbivore 
(bacteriophage ?) 

=opode epige herbivore 
(bacteriophage ?) 

*qpode epige herbivort 
(omnivore ?) 

Gasteropooe enooge CARNIVORE 

Bivalve epige symbionte 
fspece sessile. 

Bivalve epige filtreur 
espece sessile 

Bivalve epige filtreur 
espece sessile 


Lagon 
Reci f EXT. 

RfCI F EXT . 

Reci f fr. 

Recif bar. 

Lagon 

Lagon 
Lagon 


40.0 
34.09 

23.5? 

22.54 

70.97 
124 - . 33 

86.90 

103.33 


sfrsyat-.swasiSflSMS-^si 


1 <.sv 




A ■■'. .•'j'-V'.^j W 


1 *n 


k, :-i-.-.«a^v^s:j 


~3 f,\\ 




JftS'SISafflHSHSI 


1 <".(!* 




IKUUIUUilJiUfl 


1 S*l 




S ££HM 


n ti\ \ 




( 9SH9B 


1 ?a\ 




HBM 


1 1 7t 




■• 


100% 



FIGURE 3: Classement des especes etudiees, par ordre decroissant 
des vitesses de croissance, avec indication du milieu 
et du mode de vie et mention des L»o. 



19 

La figure 3 donne un classement des especes etudiees, par 
ordre decroissant des vitesses de croissance, avec, pour chacune 
d'elle, indication de son habitat et de son mode de vie; elle rappel- 
le en outre les valeurs de Loa . Quant au tableau E, il classe les 
memes especes par ordre decroissant des valeurs de production, 
en envisageant successivement la production totale (en reference 
a 1'aire du complexe recifo-lagunaire prospecte), la production 
par unite de surface (en reference a l'aire colonisee par l'espece) 
et le rapport P/B. 



PRODUCTION TOTALE 
( tonnes ) 


PRODUCTION PAR HECTARE 
(kg) 


production/biomasse 


1 Caidium fAacjum 


3500 


1 


Ca/idium /lagum 


460 




1 LiOAa/ila oHuelata 


1,59 


2 7 /lidac na maxima 


92 


2 


d/iQAaiia oi.ue.tata 


31 




2 Ca/idium //lagum 


1,00 


3 A/ica ventiicoAa 


49 


3 


7/iidacna maxima 


9 


2 


3 Neiita pticata 


0,86 


4 Chama iottoma 


16 


4 


flit/ia mitia 


6 


5 


4 7 ' ectaiiiLA g/iandinatui 


0, 36 


5 £ iota 1 ia oC.ue.lata 


4,6 


5 


A/tca ventiicota 


4 


9 


5 Chama iottoma 


0,21 


6 7 ectaaiu 4 qiandina 


tu-i 2,2 


6 


Chama loitoma 


1 


6 


6 flit/ia mit/ia 


0,19 


7 flit /to mit/ia 


0,9 


7 


7 e.cta/iiu.4 gnandina 


tuA 


4 


7 7/iidacna maxima 


0,18 


S Ne/iita plicata 


0,05 


8 


Heiita plicata 


0,01 


8 Aica uentiicota 

1 


0,14 



TABLEAU E: Classement des especes etudiees, par ordre decroissant 
des valeurs de production: production totale, production 
par unite de surface et rapport P/B. 



CONCLUSION 



En ce qui concerne les parametres de croissance, en Polynesie 
francaise, on distingue trois groupes d'especes (RICHARD, 1982a 
1983b): le Bivalve CaA.cLium ptagum (a croissance rapide), les Gastero- 
podes recifaux (a croissance relativement lente), et, enfin, les 



20 

Bivalves epiges sessiles des lagons d'atolls ferities (a croissance 
tres lente). Chama LoAtoma, objet du present travail, est caracterise 
par une vitesse de croissance tres lente, intermediate entre celles 
de T/ildaaza maxima et de Aica ventnlcoAa, especes precedemment 

etudiees dans le meme lagon (Takapoto - Tuamotu). 

En ce qui concerne la production, on separe (RICHARD, 1982a, 
1983b): les especes a tres forte production (forte production totale, 
croissance rapide, rapport P/B eleve), les especes a production 
moyenne et les especes a faible production. Avec Tildacna maxima 

et A/ica venAyilcota, Chama -Lo^toma appartient a la deuxieme categorie 
de Mollusques, quant a la production (forte biomasse, forte produc- 
tion totale, mais croissance lente et rapport P/B faible). Dans 
cette categorie, Chama -coAtoma esc l'espsce la moins productive, 
puisque les 11 millions d'individus du lagon ne produisent annuelle- 
ment que 16 tonnes de chair de Chama (soic 1,6 kg/ha). 



REFERENCES CITEES 

BERTALANFFY, von L., 1938 - A quantitative theory of organic growth. 
Human Bloloay., 10: 181-213. 

HENOCQUE, Y., 1980 - L'age du benitier, Tridacna maxima (Mollusque 

Bivalve) par examen des stries de croissance de sa coquille 

Bulletin, de -la Soclete Zoologlque de Fiance, 105, 2: 
309-312. 

RICHARD, G., 1977 - Quantitative balance and production of Tridacna 
maxima in the Takapoto lagoon (French Polynesia). 
P/ioceedlngA of. the 3id JnteAna-tLonal. Co/iaJ. 'Reef. Symposium, 
MAM, 1: 599-605. 



RICHARD, G., 1978 



Abondance et croissance de Area ventricosa 



dans le lagon de Takapoto (Tuamotu, Polynesie f rancaise) 
Hal-Lot^, 9, 1: 7-1 0. 



RICHARD, G., 1981 - A first evaluation of the findings on the growth 
and production of lagoon and reef molluscs in French 
Polynesia . 
Fouyith JnleAn.aJu.onaJ. Co/iai. HeefL Symposium, IAAN3 'LA, 2: 637 -6k1 . 

RICHARD, G., 1982 a - Mollusques lagunaires et recifaux de Polynesie 
francaise: inventaire faunistique, bionomie, bilan quanti- 
tatif, croissance, production. 
Thesie de Ooclo/iat d'Stat, fATCJS VJ , 1-2: 1-313. 



RICHARD, G., 1982 b - Bilan quantitatif et premieres donnees de 
production de Cardium fragum (Mollusca, Bivalvia) dans 
le lagon de ArJAA. 
ftalacologla, 22, 1-2: 3^7-352. 



21 

RICHARD, G., 1983 a - Growth and production of Chama iostoma in 
Takapoto atoll lagoon (Tuamotu - French Polynesia) :abstract 

JntojunatLonaA. SoaL&ty. f.0/1 Re.e.£ Stusiless,N3C(L, abA.,2k. 



RICHARD, G.,1983 b- Importance de la production malacologique 
les ecosystemes marins de Polynesie francaise. 
Qoiuinal de. la Soaiete deA OoeanlAteA , 77,XXXJX: 77-87. 



dans 



RICHARD, G., et 8. 5ALVAT, 1982 - Abondance et croissance de Tecta- 
rius grandinatus (Mollusca, Gastropoda) en Polynesie 



francaise. 
Malaco<LogJ,a, 22, 1-2: 



359-366. 



FIRST RECORDS OF WOOD SANDPIPER, RUFF, AND 
EURASIAN TREE SPARROW FROM THE MARSHALL ISLANDS 

by Manfred Temme 



Ornithological observations were made incidental to other research 
activities from 7 to 23 November 1977 and again 21 March to 12 April 
1978 on islets of Enewetak (Eniwetok) Atoll (11° 30'N, 162° 15'E) in 
the northern Marshall Islands. Birds also were observed during the 
Northern Marshall Island Radiological Survey from 11 October through 
21 November 1978 when 31 islets from Likiep, Wotho, Bikini, Ujelang, 
Kwajalein, Enewetak, and Ailuk Atolls, as well as Jemo and Me jit Is- 
lands, were visited for relatively short periods. During these trips 
several new sights and breeding records for the northern Marshalls 
were made. 

Wood Sandpiper ( Tringa glareola ) 

I observed a Wood Sandpiper on Aomon islet, Enewetak Atoll, 9 and 
21 November 1977, and saw and photographed one to three individuals 24 
and 25 March and 7 and 8 April 1978 (Fig. 1). The birds were noted in 
the company of Sharp-tailed Sandpipers ( Calidris acuminata ) , Golden 
Plovers ( Pluvialis dominica ) , and Ruddy Turnstones ( Arenaria interpres ) . 

Aomon was the only islet at Enewetak Atoll which had permanent 
freshwater ponds with some marginal and low emergent vegetation. These 
shallow, rain-filled ponds, created in 1972 by bulldozing off most of 
the coral sand cover and exposing the reef substrate, were scattered 
over approximately six of the islet's 40 hectares. 

Woodbury (1962), Pearson and Knudsen (1967), Carpenter et al. 
(1968), Amerson (1969), Johnson and Kienholz (1975), and Owen (1977a, b) 
did not mention any sightings of Tringa glareola on Enewetak Atoll or 
other atolls of the Marshall Islands. However, one female specimen 



^Manfred Temme, Center for Environmental Research and Services, Bowling 
Green State University, Bowling Green, Ohio 43403 USA 
Present address: Alter Horst 18, 2982 Norderney, West Germany 

Atoll Res. Bull. No. 292: 23-28, 1985. 



24 

(USNM 544196) in the National Museum of Natural History was collected 
on Runit (Yvonne) islet (Enewetak Atoll) on 8 September 1968 by the 
Smithsonian's Pacific Ocean Biological Survey Program (POBSP) (G. E. 
Watson in litt.)- The specimen was collected too late to be included 
by Amerson (1969), and the record has remained unpublished. 

The Wood Sandpiper breeds throughout northern Eurasia from Norway 
to the Bering Strait. It is a common migrant in Europe and southern 
Eurasia and winters in Africa, India, mainland southeast Asia, includ- 
ing the Greater Sundas and Philippine Islands , and Australia (Delacour 
and Mayr 1946; duPont 1971; McClure 1974; Temme 1974). 

Wood Sandpipers have been recorded from Midway Atoll (Clapp and 
Woodward 1968), and two birds were collected at Kure Atoll (Woodward 
1972). This species previously had been reported from western Micro- 
nesia. In the Marianas Baker (1971) considered it an uncommon visitor; 
in the Palau Islands, a regular visitor. 

There can only be speculation as to the origin of the birds at 
Enewetak Atoll. New breeding records exist from the Pribilof Islands 
and the Aleutians (White et_ aT . 1974); and numerous records (about a 
dozen specimens are in the USNM collection) are known from this area 
(G. E. Watson in litt.). That the birds could have been members of the 
pioneering Western Hemisphere population rather than Asian birds, which 
may migrate regularly as far as the Philippines and extreme western 
Micronesia, is suggested by the existing Alaskan and Hawaiian records. 
Therefore, the species should be considered a straggler and looked for 
elsewhere in Micronesia. 

Ruff ( Philomachus pugnax ) 

On 6 October 1978 one Ruff was observed on Kwajalein islet (Kwaja- 
lein Atoll). The bird was in its winter plumage and near Golden Plovers 
on the golf course. The bird was closely approached several times, and 
flushed to expose additional distinguishing features of this species. 
The relatively large size of the bird suggests that it was a male. 

At Enewetak islet (Enewetak Atoll) another sighting of a Ruff was 
made on 21 November 1979, when 0. W. Johnson pointed out a bird not 
familiar to him. Subsequently several photographs were obtained (Fig. 
2). The bird appeared smaller than the one seen on Kwajalein islet and 
may have been a female (Reeve) . It stayed in close company with 64 
Sharp-tailed Sandpipers. 

These two records, in addition to one collected on Enewetak (Clapp 
in litt.), are the only ones known for the northern Marshall Islands. 
Ruffs have been recorded from the Palau Islands (Owen 1977a) , and speci- 
mens have been collected at Kure Atoll (Clapp and Woodward 1968) and 
Pearl and Hermes Reef in the northwestern Hawaiian Islands (Amerson, 
Clapp, and Wirtz 1974) and at Johnston Atoll (Amerson and Shelton 1976). 



25 

The Ruff breeds throughout northern Eurasia from Denmark to the 
Bering Strait and winters in Africa, Pakistan, Burma, southeast China 
and casually in Japan, Taiwan, Philippines, Borneo, and Australia 
(A.O.U. 1957; McClure 1974; Temme 1974; King and Dickinson 1975). It 
is a straggler in the Marshall Islands. 

Eurasian Tree Sparrow ( Passer montanus) 

On 6, 7, 9, 29, and 31 October 1978 the only sparrow I observed 
on Kwajalein islet (Kwajalein Atoll) was the European Tree Sparrow. 
Several birds were seen in coconut palms near the town plaza and at 
gas tanks. Previously only the House Sparrow ( Passer domesticus ) had 
been noted (sight and call note records only) from this islet (Amerson 
1969). However Clapp (in litt.) suggests that _P. montanus has been 
newly introduced, perhaps via the Hawaii-Kwajalein-Guam Continental 
Airlines flight. Anderson (1981) saw a maximum of 20 Passer sp. in 
1977 but did not make specific identification. 

The European Tree Sparrow replaces the House Sparrow at about the 
90th meridian and is common in South Asia (McClure 1974; King and Dick- 
inson 1975). In the Philippines it is an introduced species and fre- 
quents human habitations (own obs.). It has been reported in Micronesia 
apparently only from the Marianas (Owen 1977b) . The observation of this 
introduced species constitutes the first recognized sight record for the 
northern Marshall Islands, though they may have been there a year ear- 
lier (Anderson 1981). These birds also were observed there by Clapp 
(in litt.) in 1979 (summer). 

I appreciated the assistance and logistic support of the Mid- 
Pacific Marine Laboratory (MPML) at Enewetak Atoll and the Department 
of Energy (DOE). W. B. Jackson, A. B. Amerson, G. E. Watson, and R. 
B. Clapp provided helpful assistance for finalizing the manuscript. 

Literature cited 

American Ornithologist's Union. 1957. Checklist of North American 
birds, fifth ed. Amer. Ornithol. Union, Baltimore. 

Amerson, A. B. , Jr. 1969. Ornithology of the Marshall and Gilbert 
Islands. Atoll Res. Bull. 172:1-348. 

Amerson, A. B. , Jr., R. B. Clapp, and W. 0. Wirtz, II. 1974. The 

natural history of Pearl and Hermes Reef, northwestern Hawaiian 
Islands. Atoll Res. Bull. 174:1-306. 

Amerson, A. B. , Jr. and P. C. Shelton. 1976. The natural history of 

Johnston Atoll, Central Pacific Ocean. Atoll Res. Bull. 192:1-479. 

Anderson, D. A. 1981. Observations of birds at Ujelang and other 
northern Marshall Islands atolls. Micronesica 17:198-212. 



26 

Baker, R. H. 1951. The avifauna of Micronesia, its origin, evolution 
and distribution. Univ. Kansas Mus. Nat. Hist. Publ. 3:1-359. 

Carpenter, M. L. , W. B. Jackson, and M. W. Fall. 1968. Bird popula- 
tions at Enewetak Atoll. Micronesica 4:259-307. 

Clapp , R. B. , and P. W. Woodward. 1968. New records of birds from the 
Hawaiian Leeward Islands. Proc. U. S. Nat. Mus. 124 (3640):l-39. 

Delacour, J., and E. Mayr. 1946. Birds of the Philippines. MacMillan, 
New York. 

duPont, J. E. 1971. Philippine Birds. Delaware Mus. Nat. Hist., 
Greenville. 

Johnson, 0. W. , and R. J. Kienholz. 1975. New avifaunal records from 
Eniwetok. Auk 92:592-594. 

King, B. F., and E. C. Dickinson. 1975. A field guide to the birds of 
South-East Asia. Houghton Mifflin Comp. , Boston. 

McClure, H. E. 1974. Migration and survival of the birds of Asia. 
Bangkok, Thailand, United States Army Medical component, SEATO. 

Owen, R. P. 1977a. New bird records for Micronesia and major island 
groups in Micronesia. Micronesica 13:57-63. 

Owen, R. P. 1977b. A checklist of the birds of Micronesia. Micro- 
nesica 13:65-81. 

Pearson, D. L. , and J. W. Knudsen. 1967. Avifauna records from 
Eniwetok Atoll, Marshall Islands. Condor 69:201-203. 

Temme, M. 1974. Beitrag zur Kenntnis des Vorkommens ostpalMarktischer 
Limikolen auf Mindoro, Philippinen. J. Ornithol. 117:100-104. 

White, C. M. , F. S. L. Williamson, and W. B. Emison, 1974. Tringa 

glareola - a new breeding species for North America. Auk 91:175- 
177. 

Woodbury, A. M. 1962. A review of the ecology of Eniwetok Atoll, 

Pacific Ocean. Univ. Utah, Inst. Environment. Biol. Res.:l-123. 

Woodward, P. W. 1972. The natural history of Kure Atoll, northwestern 
Hawaiian Islands. Atoll Res. Bull. 164:1-318. 



27 




Fig. 1. Wood Sandpiper Tringa glareola on Aomon islet, Enewetak 
Atoll, Marshall Islands (April 8, 1978) 

Fig. 2. Ruff Philomachus pugnax among Sharp-tailed Sandpipers 
Calidris acuminata on Enewetak islet, Enewetak Atoll 
(November 17, 1978) 




CLASSIFICATION OF EMERGENT REEF SURFACES 
by F. R. Fosberg 



Anyone who has attempted serious bibliographic work on "coral 
reefs" (sensu latissimo), or on islands, or on coastal zone features, 
has likely had a feeling of being overwhelmed by the sheer amount of 
published (and unpublished) information that has accumulated. The 
realization inevitably leads to analytical attempts to break the mass 
down into more manageable fragments or "fields". These are normally 
arranged into classifications of one sort or another. The nature of 
such analyses and classifications ordinarily is determined by and re- 
flects the particular interests and biases of those who make them. 

Sometimes, in any area of knowledge, one finds that a more or 
less satisfactory scheme already exists, and can be accepted and used 
or modified. Most often, because of differences in purposes or objec- 
tives of one's investigation, available schemes are not entirely ap- 
propriate and must be modified to suit new requirements, or new schemes 
must be worked out. This may be true at any level in the hierarchy of 
the classification of knowledge. 

The present attempt resulted from an essay on the present state 
of knowledge of the floras and vegetation of emergent reef surfaces — 
the terrestrial plant cover of relatively recent emergent reefs. 

Rough or vague arrangements of information on these may be found 
in almost any consideration of the vegetation of islands, or of a 
single island or group, less frequently in floristic works. These 
schemes usually assume a knowledge or familiarity with "the obvious" 
and often use undefined general units or local folk terms and concepts. 
They are often more or less satisfactory for the immediate area con- 
cerned, but become less os or unsatisfactory when extended, generalized, 
or adapted to other regions. The scheme here presented evolved because 
none of the familiar ones was broad enough, precise enough, or exactly 
served to facilitate understanding of the botany of emergent reef sur- 
faces. 



Atoll Res. Bull. No. 292: 29-36, 1985. 



30 

Even the term "emergent reef surface" presents difficulties. Ob- 
viously it means the surface of a reef that is above high tide level. 
But what about fossil reefs? In a sense, all even slightly emergent 
reefs, even elevated beach-rock, are fossil reefs. How far back should 
one go geologically? Ideally, emergent reef surfaces should be sur- 
faces that have never been buried by massive sedimentary deposits or by 
volcanic ejecta — either ash or lava — though they may bear accumulations 
of sand, gravel, boulders or other reef-derived material. They are 
calcareous in composition except for small amounts of pumice or other 
water-borne or wind-borne substances. This means that the plant cover 
of these surfaces derives from successful plant colonizations over the 
period since the emergence of the reefs, by lowering of sea-level, 
tectonic uplift, or accumulation of water-carried calcareous material 
(bars, spits, or storm ridges). 

This definition of emergent reef surfaces excludes most geologi- 
cally old or ancient reefs, as their exposed surfaces have mostly been 
formed by erosion of overlying non-reef material. Usually their lime- 
stones are of quite different character than those of the more recent 
reefs that have never been buried. Their floras bear little or no re- 
lation to the strand flora or enriched strand flora of modern emergent 
reefs. In general, fossil reefs of Quaternary to Recent age provide 
emergent reef surfaces. Some are classed as Plio-Pleistocene , and in 
rare instances as Miocene (e.g. the Barrigada Limestone of Guam, and 
limestones on Mangaia, Cook Islands). Of course, there has been more 
or less chemical erosion, and even abrasion, on all or most such sur- 
faces. 

The following scheme will provide a basis or framework for corre- 
lation of the plant species and vegetation with appropriate variations 
in the emergent reef habitat. It is arranged in outline form, but 
with sufficient descriptive comment to make the distinctions and rela- 
tionships clear. The principal basis of the classification is topo- 
graphic and locational, but these factors are strongly correlated with 
degree or lack of induration, and degree of chemical erosion and con- 
sequent roughness or ruggedness of the surface. Almost all of the 
units listed show differing facies in areas of greater or less rain- 
fall. Solution of the limestone and consequent change or degradation 
of the surface may be more active in areas of greater rainfall, as may 
the effects of plant roots and humus. Different burrowing animals may 
affect the surfaces in different ways, and their distribution may be 
influenced by the climate. The rainfall factor strongly influences the 
vegetation and flora directly, and the more abundant soil formation in 
wet areas produces perceptible effects on the reef surface. Salt spray, 
too, affects the nature of the erosion of coral limestone, in a way not 
fully understood, at least by me. This will be described at an approp- 
riate place in the classification. 

(1) A convenient primary subdivision of emergent reef surfaces is into 
(1.1) oceanic , that is, formed on or around islands which have never 
had any connection or close proximity to continental or major island 
land-masses, vs. (1.2) continental, formed on the shores of continental 



31 

or large island land-masses. These categories are appropriate for bio- 
logical purposes, because of the differences in complexity of oceanic 
vs. continental biotas (see 1.2 below). 

(2) Oceanic emergent reefs are either (2.1) atolls and table reefs , 
not closely associated with non-calcareous islands, or (2.2) fringing 
and barrier reefs , formed around higher, usually volcanic islands. 

(3) Atolls and table reef surfaces are either essentially at (3.1) 
sea-level or (3.2) relatively uplifted. 

(4) Essentially sea-level reefs may be of (4.1) loose, unconsolidated 
material, sand, gravel or boulders, or of (4.2) indurated such material 
or in-place reef structure; either kind may bear temporary local accumu- 
lations of loose sand (dunes) or gravel storm-ridges which may reach 
over 3 m, rarely considerably more. 

(4.1) Unlithif ied reef surfaces, including sand cays , islets and bars, 
are loose accumulations of sand-size or larger foraminiferal tests and 
fragments or entire skeletons of other calcareous animals and plants. 
In areas where the ocean is generally only slightly or moderately tur- 
bulent there may be, especially on lagoon margins, deposits of precipi- 
tated or triturated silt-size material. This material is usually blown 
away by wind but may be held by algal crusts or evaporite salt crusts. 
Cays of loose material tend to be changed frequently by storms and 
wave action, at least until they become well-stabilized by vegetation 
or their margins become protected by intertidal beach-rock formation. 

(4.2) Lithified atoll islets and table reefs have at least parts of 
their surface of a cemented reef-conglomerate or a lime-sandstone 
platform, or of bedded atoll phosphate rock ( Jemo soil ). There is 
little agreement as to the circumstances under which such lithifica- 
tion takes place, but the physiography frequently suggests that it 
happened during previously higher post-glacial sea-levels of 2 or 3.5 
m above present. The phosphatic lithif ication is associated with 
present or past Pisonia grand is forests and roosting or nesting of 
sea-birds. Lithified surfaces on these very low islands are usually 
flat, but older ones may be rough or pitted by chemical erosion. 

(5) Elevated reefs may be either (5.1) slightly raised (4-8 m) or (5.2) 
substantially more so, and almost always at least mostly indurated, but 
often with some perched loose material, storm- or wind-deposited. Sur- 
faces may be either relatively flat, or dissected; older strongly 
elevated reefs may be eroded into (5.3) karst topography, but karsts 
also may be cut into much older limestones of other than reef origin. 
This distribution may not be easy to establish. 

(5.1) Slightly elevated atolls and table reefs . Usually partly of 
sand or gravel, but at least with a core or an extensive platform of 
conglomerate or "reef-rock" that extends locally to more than 4 m ele- 
vation. The lithified surface is flat and covered by loose deposits, 
or bare and either a flat "pavement" (platin) or a pitted, pinnacled, 



32 

or intricately dissected or eroded into a sharp "fretwork" ( champignon ) , 
or less so ( pave ) . These parenthetical terms are Creole, used in the 
Western Indian Ocean islands, especially Aldabra. A Polynesian term 
for dissected slightly elevated surfaces is feo . Caribbean terms for 
sharply dissected surfaces are "dog-tooth" and "iron-shore". The very 
sharply and finely dissected or "fretwork" facies of this surface seems 
associated in someway with proximity to salt-water, perhaps exposure to 
salt spray. Simply pinnacled facies may be found more inland, out of 
reach of heavy spray. How sea-water, which is said to be super- 
saturated with calcium carbonate, can cause or accentuate chemical 
erosion of limestone is not clear. Long-term experimental work on 
this problem would be desirable. 

(5.2) Elevated flat-topped reefs (10-200 m or more) not closely associ- 
ated with high islands or continental shores are not numerous, and the 
surfaces of most of them have been destroyed or completely altered by 
phosphate-mining. On Nauru, for example, an artificial mini-karst or 
deeply (to 10 m) pitted-pinnacled new surface has been produced. Natural 
pinnacled surfaces exist on Henderson Island, nearly undisturbed. On 
Henderson, also, are relatively smooth surfaces of loose material or 
soil. Such islands are generally surrounded by cliffs, with or without 

a narrow flat coastal strip or shelf at or just above sea-level at base. 
The cliffs may be vertical or very steep, with, in places, ledges or 
caves. They may be undercut, intertidally. 

(5.3) True karsts are a third sort, more often near larger islands (or 
on continental shelves). They are very rugged, with sharp peaks and 
ridges, steep rough slopes, and are often deeply undercut at base. 
Without careful geological observation it is not always certain whether 
they are cut in elevated reefs or in ancient limestones of other than 
reef origin. Good examples are the southern islands of the Palau group 
(except Angaur and Peliliu) and the Lau Group of Fiji (possibly contin- 
ental rather than oceanic). 

(2.2) Reefs around high islands . The difference between these and 
those listed above is important biologically because of the enrichment 
of the biota due to the proximity to the more diverse biota of the 
high islands. As with the reefs not associated with high islands, 
those in this category may be divided into (6.1) essentially sea-level 
and (6.2) significantly elevated above sea-level. 

(6.1) Sea- level- fringing and barrier reef emergent surfaces are similar 
in most ways to those of atoll islets. The islets on barrier reefs 
are, in fact, scarcely distinguishable from those of atolls except for 
the proximity of a high island. Emergent fringing reefs may present a 
smooth, abraded surface, or a variously pitted one. Non-calcareous 
earth may wash or blow down onto the fringing reef surface, making pos- 
sible the growth of more species of plants. 

(6.2) Elevated reefs surrounding high islands may be (7.1) slightly 
elevated or (7.2) strongly elevated, as are those in open sea. 



33 

(7.1) Slightly elevated reefs , either barrier reef islets or fringing 
reefs. These present a series of surface features similar to those 
described above for slightly elevated atolls and table reefs (see 5.1). 

(7. 2) Elevated reefs - peripheral and terraces on slopes . These are 
frequently deeply dissected, sometimes labryrinthine. The "makatea" 
surrounding some South Pacific islands (e.g. Austral Islands) is an 
example . 

(1.2) Continental reefs , lining coasts of continents and continental 
islands, comprise a series of categories parallel to that outlined 
above for oceanic island reefs but, by their geographic positions, 
carrying richer floras and more complex vegetation. Descriptions of 
such categories need not be repeated here. Continental reef surfaces 
are numerous in the Caribbean and western Pacific regions. 

The above scheme is summarized in the accompanying diagram. 

In a review of this paper the suggestion was made that the classi- 
fication here proposed be related to previous classifications of the 
same phenomena. The literature on reef-classification is enormous and 
I am familiar with much, but by no means all, of it. By far the greater 
part of it concerns submerged reef features and the processes that pro- 
duce them, the preoccupation of most students of coral reefs. Emerged 
features are mentioned incidentally, or in relation to such features as 
soils, geology, geomorphology , and land ecology. Descriptions usually 
apply to specific examples — atolls, islands, or localities. Many of 
the descriptive terms, and the phenomena referred to, are of wider oc- 
currence or application, and are useful for general descriptive purposes 
while others are too limited or specific to be generally applicable. 
Many of the terms and features used or described in the present attempt 
are from one or more such papers, but nowhere have I seen a comprehensive 
description or classification of emergent reef surface phenomena. The 
one that comes closest is in the sections on Terrain and on Islets, as 
well as scattered through the text, of my volume on Military Geography 
of the Northern Marshall Islands (1956). This includes and describes 
many of the surfaces treated in the present paper, and furnishes much 
of its substance. However, it is limited to the surfaces found on the 
sea-level atolls of the Northern Marshall Islands and is not organized 
into a classification. It does not provide the inclusive array of 
emergent reef surfaces that exist and is not thoroughly applicable to 
all known atolls and barrier reef islets. Elevated surfaces are not 
treated at all. 

I have read and been impressed by Stoddart's pungent discussion 
of the confused state of reef terminology in Coral Reef Research Methods 
(Stoddart and Johannes, 1978). I hope I have not contributed further 
to the confusion he described. It is noticeable that, after this per- 
tinent discussion, the volume does not offer a terminology which would 
help avoid the difficulties pointed out, nor does it include any attempt 
toward a classification or orientation in the kinds of reefs or reef 
surfaces. 



34 



Emergent Reefs 
1 



Oceanic 
1.1 



Continental 
1.2 



Atolls 

Table Reefs 

2.1 



Fringing and 

Barrier Reefs 

of High Islands 

2.2 



At Sea Level 
3.1 



Elevated 
3.2 



At Sea Level 
6.1 



Elevated 
6.2 



Slightly Elevated 

4-8 M 

5.1 






Elevated 

10-200 M 

5.2 



True Karst 
5.3 



Slightly Elevated 
7.1 



ZZ] 

Elevated 

Makatea 

7.2 



Unconsolidated 
4.1 



n_ 



Lithified 
4.2 



Fig. 1. Diagram of Classification of Emerged Reef Surfaces 



35 
REFERENCES CITED 

[Fosberg, F. R.U. 1956. Military Geography of the Northern Marshall 
Islands. 320 pp. pub. by Office of the Engineer, Hq. U.S. Army 
Forces Far East, L Tokyo] . (Unclassified). 

Stoddart, D. R. and R. E. Johannes, eds. 1978. Coral Reefs: Research 
Methods. 581 pp. UNESCO, Paris. 



BOTANICAL VISITS TO KRAKATAU IN 1958 AND 1963 
by F. R. Fosberg 



Twice, on December 20, 1958 and December 9, 1963, while visiting 
Java for UNESCO Humid Tropics Program activities, I had the privilege 
of seeing Krakatau Island volcano. The recent celebration (1983) of 
the 100th anniversary of the world's greatest explosion in recorded 
history, and the publication of the complete collected writings about 
the Krakatau eruption (reviewed briefly below) , brought to mind ob- 
servations I made on my visits, 25 and 22 years back. 

Brief notes on the conditions of Anak Krakatau (Krakatau' s Baby), 
the new cone that appeared on the site of the center of the former 
large volcanic cone may be of some interest. Comparison of the list 
of plants seen and collected on Anak in 1963 with lists recorded be- 
fore and after may contribute to understanding of the ability of plants 
to cross water barriers and to colonize new volcanic substrata. A 
valuable addition to this account is a list of the collections made, 
at my suggestions, on Anak Krakatau in August 1971, by Professor 
Mildred Mathias and her colleagues, Phung Trung Ngan and W. Soegeng 
Reksodihardjo. Her specimens are at U.C. Los Angeles. Mine are in 
the U.S. National Herbarium, Smithsonian Institution, Washington, D.C. 

On December 20, 1958, aboard the Indonesian oceanographic research 
vessel Samudera, a turn was made around Anak Krakatau, as close in as 
was prudent, during a period of pulsating explosive eruptions. My 
attention was so held by the excitement of the eruption that I did not 
write much down. I merely noted several Casuarina shrubs on the south 
slopes. A brief, edited, account of my 1958 and 1963 notes follows. 

The three remnants of the original volcano, very steep and rugged, 
are arranged in a broken circle, densely wooded to their summits ex- 
cept for vertical cliffs. Time was not available to land on them. The 
group reminded one of Maug, in the northern Marianas, except on a much 
larger scale and densely wooded. 



Atoll Res. Bull. No. 292: 39-48, 1985. 



40 

Anak, during our circuit in 1958, was in a state of continuous 
pulsating activity. Explosions occurred every few minutes, throwing 
ash and rocks to considerable heights. From the rim of a large crater 
smooth slopes of dark brown ash and cinders ended in low wave-cut 
cliffs, except at one end where the slope reached beach level. Ap- 
parently some coral debris may have been cast up here, as the top of 
the beach, otherwise black, is light colored. Here a small patch of 
Casuarina had reached tree size. Three or four similar trees were seen 
scattered above on the slopes. Otherwise no vegetation was seen through 
vinoculars. 

The forest on the three outer, older islands was luxuriant to the 
tops, except on the cliffs. Casuarina is an important compenent, mostly 
in patches. Terminalia catappa was perhaps the most important tree in 
area, covering much of the lower slopes. However, there were a good 
many other species and the vegetation was generally a dense mixed forest. 
No grass was visible except on the steep inner wall. 

On our visit in December 1963, again under UNESCO auspices, the 
volcano was quiet, though producing much steam and sulphur dioxide. 
We were able to land at the northeast corner, where the same clump of 
Casuarina , seen in 1958, was now much taller. The steam, in great 
white billows, was mainly coming from the low south side of the crater. 

From the sea, before landing, we could examine the mountain with 
binoculars. The crater was about one third or half the width of the 
island. The south side had been undercut and slumped, exposing very 
clear bedding of the ash. From fairly close, a scattering of small 
plants, grass tufts, could be seen on the slopes above the Casuarina 
clump, and one fair-sized Casuarina some way up toward the crater. The 
lowest point on the crater rim appears to be where a lava flow has come 
forth and down the south slope, spreading out fan-wise at the base. 
The flat ground around the Casuarina clump at the northeast corner is 
grassy, the grass extending westward along the north shore, to a smaller 
sparse patch of Casuarina , mostly appearing dead. 

Ashore the flat area of grassy vegetation contained about 25 
species, some of them only represented by one or more germinated seeds. 
They are listed below. Only the abundant sterile tufts of Saccharum 
spontaneum and a few scattered Casuarina ascended part way up the slopes. 
Land-crab holes were seen on the slope as much as 300 m from the sea, 
and dead grasshoppers even father up. 

Huge splatter-bombs have been thrown at least 2/3 the way from the 
crater to the sea, smaller ones even farther. Nearer the top they prac- 
tically cover the slope, along with scattered pieces of a dense porphyry, 
several colors of scoria, some pumice, and a few enormous fusiform bombs. 

Inside the crater were white bedding, and a complex of cones and 
vents, brightly stained with sulphur. The congealed flow seen from the 
sea runs along the area of cones and spills over the rim to the south. 
Steam and sulphur dioxide discouraged much lingering at the top. We 
climbed only the north slope and descended to the northwest corner. 



41 

On the flats of cinders and volcanic sand were low thickets of 
Casuarina and patches of tall Saccharum spontaneum , sod of Ischaemum 
and mats of Ipomoea pes-caprae and Canavalia rosea , which seemed to have 
repeatedly been almost killed, probably by fumes. Some of the Casuarina 
was dead or almost so, as well as the Morinda and Calophyllum . Cassytha 
was mostly dead. Ischaemum leaves were dead but the stems were still 
green. All species seen living were collected, but there were also 
scattered wave-cast seeds that had not germinated. Time was too limited 
to gather these. 

The following list includes the species recorded by J. van Borssum 
Waalkes in 1960 (Ann. Bogor. 4:5-64) of all species collected or ob- 
served on Anak by himself or other visitors before all the vegetation 
was destroyed by the 1952 eruption. Pre-1952 records are indicated by 
W, van Borssum Waalkes collections by W with his collection numbers; 
Fosberg records in 1963, with collection numbers, by FRF; and 1971 
records by Mathias et al. by MEM plus collection numbers. Field notes 
and comments of interest are included in parentheses after the approp- 
riate collection numbers. Adjustment of the nomenclature has been made 
where discrepancies between the lists are found. 

Lygodium f lexuosum (L. ) SW. 

MEM 25 

Nephrolepis falcata (Cav. ) C. Chr. 
MEM 28 

Nephrolepis hirsutula (Forst . f.) Presl 
W 1068 (on slopes to 50 m) 

Nephrolepis radicans (Burm. f.) Kuhn 
MEM 29 

Pityrogramma calomelanos (L.) Link 

W 1069 (on slopes to 50 m, fair numbers seen in 1951) , MEM 32 

Cycas circinalis L. ( Cycas rumphii Miq.) 
W (one plant seen) 

Pandanus tectorius Park. 

W, FRF 44549 (only a few plants seen) , MEM 30 

Imperata cylindrica Beauv. 

W 1081 (small area, only, on E side), FRF 44555 (rare, one tiny 
patch seen) , MEM 21 

Ischaemum muticum L. 

W 1083 (fairly large number seen in 1951), FRF 44551 (very 
common, forming loose sod locally), MEM 24 

Saccharum spontaneum L. 

W 1075, FRF 44556 (commonest plant on island, mature ones on 
flat, smaller ones on cinder slopes), MEM 35 



42 

Spinifex littoreus (Burm. f.) Merr. 

W 1070, FRF 44547 (occasional on open beach), MEM 37 

Thuarea involuta (Forst. f.) R. & S. 
W 1074, MEM 39 

Cyperus javanicus Houtt. 

W 1071, FRF 44552 (rare, one tuft seen), MEM 10 

Fimbristylis cymosa R. Br. 

FRF 44553 (rare, one tuft seen) 

Remirea maritima Aubl. 

FRF 44550 (very local in sand) 

Cocos nucifera L. 

W (germinating nuts seen, probably both washed ashore and planted 
by man) , MEM 8 

Nypa fruticans Wurmb. 

W 

Musa paradisiaca L. 

W 

Eulophia pulchra (Thou.) Lindl. ( Eulophis macrostachya Lindl.) 
MEM 14 

Casuarina equisetifolia L. 

W 1082, FRF 44562 (common, spreading up slopes; seen in 1958), 
MEM 6 

Piper aduncum L. 
MEM 31 

Ximenia americana L. 

MEM 42 

Ficus septica Burm. f. 

MEM 18 (observed only) 

Ficus fulva Reinw. ex Bl. 

MEM 17 

Cassytha filiformis L. 

W 1075 (fairly large numbers seen in 1951), FRF 44545 (abundant), 
MEM 5 (evidently common, associated with several host plants) 

Hernandia sonora L. 

W (very young seedlings, only) 

Albizia retusa Benth. 
MEM 1 



43 

Canavalia rosea (Sw.) DC. 

W 1077, FRF 44544 (very common), MEM 4 

Derris trifoliata Lour. 

W 

Desmodium umbellatum (L.) DC. 

W 1078, FRF 44559 (occasional), MEM 11 

Erythrina variegata L. 

W (seen earlier but not in 1951), FRF 44560 (rare, one plant 
seen), MEM 13 

Phaseolus sp. 

W 

Pongamia pinnata (L.) Pierre 

W, FRF 44565 (occasional) , MEM 33 

Vigna marina (Burm. ) Merr. 
W, MEM 40 

Murray a exotica L. 
W 

Xylocarpus granatum Koen. 
W 

Antidesma sp. 
W 

Euphorbia chamissonis (Kl. & Gke.) Boiss. 
MEM 16 

Dodonaea viscosa L. 

MEM 12 (seedling) 

Colubrina asiatica (L.) Brongn. 
MEM 9 

Cissus repens Lam. 
W 

Hibiscus tiliaceus L. 

FRF 44564 (rare), MEM 20 

Calophyllum inophyllum L. 

W 1080, FRF 44566 (occasional, mostly dead), MEM 3 (seedling) 

Barringtonia asiatica (L. ) Kurz 

W (seen earlier but not seen in 1951), FRF 44567 (rare, seedlings 
only) , MEM 2 (seedling) 



44 



Rhizophora mucronata var. stylosa (Griff.) Schimper ( Rhizophora 
stylosa Griff.) 
W 

Terminalia catappa L. 

W 1083, FRF 44563 (occasional), MEM 38 

Melastoma malabathricum L. (Melastoma polyanthum Bl.) 
MEM 26 — ~ 

Cerbera manghas L. 
W, MEM 7 

Ipomoea littoralis Bl. (as Ipomoea gracilis R. Br.) 
FRF 44557 (rare, one plant in thicket), MEM 22 

Ipomoea pes-caprae ssp. brasiliensis (L. ) v. Ooststr. 
W 1076, FRF 44548 (common), MEM 23 

Clerodendrum inerme (L.) Gaertn. 

FRF 44561 (rare, one plant seen) 

Premna obtusifolia R. Br. 

FRF 44558 (occasional), MEM 34 

Guettarda speciosa L. 
MEM 19 

Morinda citrifolia L. 

W 1067, FRF 44546 (occasional), MEM 27 

Scaevola sericea L. 

W 1079 (common, but much eaten by grasshoppers), FRF 44568 
(occasional) , MEM 38 

Chromalaena odorata (L. ) King & Rob. ( Eupatorium odoratum L. 
FRF 44554 (occasional), MEM 15 

Wollastonia biflora (L. ) DC. ( Wedelia biflora (L. ) DC.) 

MEM 41 

******************* 

Since this paper was written an excellent article on Krakatau, by 
Ian W. B. Thornton (Ambio 13:216-225, 1984), has come to my attention. 
This artical summarizes what was on record as to the recolonizations 
over the 100 years following the great explosion in 1883, and adds ob- 
servations made during the Hull University Expedition in 1983, by J. R. 
Flenley and K. Richards, as well as by the author himself, on three 
visits. 

It may seem superfluous to publish the observations made in 1958, 
1963 and 1971, after the excellent Thornton account, However, the 



45 

information offered here was not available to Thornton, and partially 
fills in the period between two total or almost total sterilizations of 
Anak Krakatau by major eruptions (1952 and 1972). It is unfortunate 
that a collection made by Kostermans at some time between 1952 and 1963 
could not have been included. According to Kostermans (personal com- 
munication in 1964) the specimens collected were incorporated into the 
Bogor herbarium and no list of them was preserved. Retrieval of the 
record of species with any degree of completeness would be impractical. 
It is hoped that the results of the series of investigations, 1982-1984, 
mentioned by Thornton, at least so far as they concern Anak Krakatau 
plants, may be assembled and studied with an aim to better understand 
the processes of dispersal, establishment, increase, decrease, and dis- 
appearance, chance and probability, in relation to succession. A more 
balanced comparison with the detailed studies of Surtsey (Iceland) may 
then be possible. 

Thornton (p. 224) quotes Tagawa (1983) as listing Nephrolepis 
tomentosa v.A.v.R. as one of the dominant pioneers on lava flows. I 
have listed three other Nephrolepis species as present, one in 1951, 
the others in 1971. I have not been able to consult a specimen of N_. 
tomentosa from anywhere, nor have I seen the Anak Krakatau collections 
of any of the four. So I cannot comment on the possibility that all 
of these records may be different identifications of the same species 
in this extremely difficult genus. Also quoted as occurring on ash- 
covered lava flows is Melastoma af finis . This may very probably be 
the same as what I have recorded as M. malabathricum L. sensu lato. 

A definitive study of the flora and successional vegetation of 
Anak should include comparison of all of the actual specimens previously 
gathered there. In this way, only, can we be sure of how many and which 
species we are dealing with. 

Appended are several lists that indicate changes in the Anak flora 
since the appearance of this new volcanic cone in 1930. 



46 



Species present previous to 1951: 

Nephrolepis hirsutula 
Hernandia sonora 
Pityrogramma calomelanos 
Thuarea involuta 
Cycas circinalis 
Antidesma sp. 
Cerbera manghas 
Cissus repens 
Derris trifoliata 
Murraya exotica 
Musa paradisiaca 
Nypa fruticans 
Phaseolus sp. 
Xylocarpus granatum 
Vigna marina 

Species present in 1963: 

Canavalia rosea 

Cassytha filiformis 

Morinda citrifolia 

Spinifex littoreus 

Ipomoea pes-caprae ssp. brasiliensis 

Pandanus tectorius 

Remirea maritima 

Ischaemum muticum 

Cyperus javanicus 

Fimbristylis cymosa 

Chromalaena odorata 

Imperata cylindrica 

Saccharum spontaneum L. var. klagha 

Ipomoea littoralis 

Premna obtusifolia 

Desmodium umbellatum 

Erythrina variegata 

Clerodendrum inerme 

Casuarina equisetifolia 

Terminalia catappa 

Hibiscus tiliaceus 

Pongamia pinnata 

Calophyllum inophyllum 

Barringtonia asiatica 

Scaevola sericea 

Species present in 1963 but not found in 1971: 

Remirea maritima 
Fimbristylis cymosa 
Clerodendrum inerme 



47 



Species present in 1971, but not found in 1963: 

Albizia retusa 
Cerbera manghas 
Cocos nucifera 
Colubrina asiatica 
Dodonaea viscosa 
Eulophia pulchra 
Euphorbia chamissonis 
Ficus fulva 
Ficus septica 
Guettarda speciosa 
Lygodium flexuosum 
Melastoma malabathricum 
Nephrolepis falcata 
Nephrolepis radicans 
Piper aduncum 



CHECKLIST OF THE HERPETOFAUNA OF THE MASCARENE ISLANDS 
by D. D. Tirvengadum* and R. Bour** 



ABSTRACT 

An up to date checklist of the reptiles and emphibians of Mauritius, 
Rodrigues, Round Island and Reunion is presented. Maps showing subfossil 
collecting sites and localities where living specimens had been reported 
by old settlers, sailors, or travellers are also included. 

RESUME 

Une liste a jour des reptiles et amphibiens de l'lle Maurice ainsi 
que de Rodrigues, de l'lle Ronde et de la Reunion est presentee, et 
egalement des cartes precisant les sites de collecte de subfossiles et 
les endroits ou ont ete reperes des specimens vivants par d'anciens 
colons, des marins, ou des voyageurs. 

The small Mascarene islands, Mauritius, Rodrigues, Reunion and par- 
ticularly Round Island, an offshore islet lying northeast of Mauritius, 
supported a poor but interesting and peculiar reptile fauna. Over the 
past three centuries with the advent of man followed by large settlement, 
the reptile populations of these islands, the endemic elements in par- 
ticular, have suffered considerably from degradation of their natural 
habitats and from the depredation of domestic animals. 



* Previously Director, Mauritius Institute, Port-Louis, Mauritius 
Present address: Laboratoire de Phanerogamie, Museum National 
d'Histoire Naturelle, 16, rue Buff on, 75005 Paris, France 

** Laboratoire des Reptiles et Amphibiens, Museum National d'Histoire 
Naturelle, 25, rue Cuvier, 75005 Paris, France 

Atoll Res. Bull. No. 292: 49-60, 1985. 



50 

Reports from various sources in Mauritius and Reunion and a survey 
carried out by the authors in Rodrigues and on Round Island in 1980 re- 
veal that the composition of the reptile fauna have been greatly 
modified: impoverishment of the previously abundant endemic species, 
precarious existence on Round Island of the most remarkable representa- 
tives of the reptile fauna of the Mascarenes. There has also been a 
depletion of the introduced species (tortoises, frogs, agamids, snakes). 

A reappraisal of all existing genera and species hitherto described 
permits the publication of a complete list of extant and extinct rep- 
tiles and amphibians of the Mascarene islands. A list of amphibians 
and reptiles erroneously cited in literature or accidentally met with in 
the Mascarene islands is also given. 

ACKNOWLEDGEMENTS 

This paper was prepared in 1980 and submitted for publication in 
the 1981 issue of the Mauritius Institute Bulletin. That issue was 
however cancelled because of technical reasons. We are therefore 
grateful to the Editors of the Atoll Research Bulletin for having ac- 
cepted to publish this manuscript in its present form. 

We are also grateful to Dr. E. N. Arnold (British Museum, Natural 
History, London) and to Dr. F. Moutou (Ecole Veterinaire, Maisons-Alfort , 
France) for helpful discussion. 

Mr. Roger Bour helped in setting up a permanent showcase on the 
reptile fauna of Mauritius and the neighbouring islands in the Natural 
History Museum of the Institute on the occasion of the centenary of the 
Mauritius Institute in November 1980. 






co 

<D 
■H 
O 
QJ 

a 

CO 



co 

•H 



w 
-J 

< 



"•CD 

Pi 



05 

QJ 

3 
00 
■H 

u 

X) 

o 

OS 



x 

c ca 

3 r-l 
O CO 

PS I— I 



u 
3 

CO 



<** 



4< 



o o o 



■n 



* * 



w 



oo 



o 



* -X O O 













p 






























CO 

3 
















o 






























•H 


y~~** 














CO 


•— •* 




























4-1 


CO 












^-s 


c 


5-1 




























•r-l 


3 












CO 


•H 


QJ 




























5-1 


•H 












c 


> 


X 




























3 


4-) 












•H 




4-> 




























CO 


•H 












X) 


°3 


C 




























S 


14 












J-l 




:p 






























3 












CO 


P 


a 




























«N 


CO 


^\ 










- 1- 1 


O 














C 
















c 


5H 


CO 










CO 


CO 














O 














,^-s 


•r-l 


^ ' 


3 










QJ 


C 


ca 












CO 














p 


o 




•H 










a 


•iH 
> 


c 
ca 


^ 






CO 

c 

QJ 




C 

•H 

> 














o 

•r-l 
P 


& 


CO 

C 
OJ 


4-1 

•H 
!-4 










■H 


ca 


4-1 


CO 




QJ 


4-> 


















3 


CO 


4-> 


D 










■H 





•H 


c 




^ 


M 




o3 










•— N 




^QJ 


— 


S-l 


ca 










C 


•r-1 


5-4 


■H 




QJ 


JJ 














•H 




PS 


5-1 


OJ 


s 










o 


P 


3 


XJ 




XI 






p 










P 






OJ 


s: 


**-^ 










4-) 


£ 


ca 


M 




u 

















o 






P 














3 


E 


ca 


•— >* 




ca 


•~\ 


CO 




„ 


5-4 




CO 




CO 


c 


•H 


p 




• N 


^— "\ 




£ 


n 




•i—i 


CO 


QJ 


o 


4-) 


p 




* 


0) 




P 


>^ 


p 


3 


^! 


o 




c 


5-4 




£ 


^— s 


CO 


T3 


ca 


■H 


J-l 


•H 




CO 


OO 




•H 


CO 


0) 


O 


CO 


CO 




•H 


QJ 






CO 


CD 


H 


00 


c 


QJ 


> 




p 


§ 




> 


5-i 


4-) 


s — ' 


OJ 


p 




X3 


•H 




■H 




3 


a 


ca 


QJ 


5 


X 




•~N 


Q) 






CJ 


5-1 




X 


•r-l 




3 


> 




•r-l 


•H 


5-1 




N 


1—1 


1—1 


•H 


X) 


4-) 


1—1 




II 




OJ 


CO 


E 


D> 




ca 


3 




c 


■H 


3 




•r-l 


CO 


J-l 


•H 


•H 


J-l 


J-l 


3 






ca 


S 


•H 


•H 






Q 


O 




o 


H 


ca 


•H 


1—1 


00 


B 


s 


C 


CO 


QJ 


O 


>^ 




4-) 




CO 


U 


co 




• — ' 


* — y 




J-l 


QJ 


CO 


•H 




CO 




O 


P 


S 


ca 


CO 


X 


ca 


ca 


p 


00 


• H 










3 


•'—) 


o 


M 


p 








4-) 


V QJ 






5-1 


CO 


p 


4.) 


OJ 




w 




M 


CO 




5 


O 


X> 


•H 


QJ 


ca 


ca 


ca 


s 


■i-l 


•H 


■H 


O 


5-1 


5-1 


CO 


J-l 


•H 


CO 




o 


■H 




X 


•H 


ca 


QJ 


o 


o 


p 


aj 


rJ 


3 


5-1 




o 


o 


4J 


■H 


3 


.—1 




1— 1 


i— 1 








a 


CM 


J-l 


•H 


•H 


ca 


p 




CO 


QJ 


CO 






o 


E 


ca 


3 




o 


cC 




CO 


CO 




i— 1 


U 


c 


c 


•H 


X) 




QJ 


X 


4-> 


CO 


CO 


01 


QJ 


-S 


OO 




o 


X) 




3 


3 


ii 


QJ 




o 


o 


T3 


5-1 


CO 


X 


4-> 


cfl 


4-J 


4-> 


ex 


X) 


•H 




•1-1 


M 




M 


.p 




4J 




XI 


X 


QJ 


ca 


CO 


E 


R 


OJ 


CO 


CO 


X 


C 


E 


OJ 




CO 


ca 




ca 


& 






CO 


u 


J-l 


CX 


5 


OO 


•H 


QJ 


p 


g 


p 


QJ 


•H 


■H 


CO 




M 


a, 




x 


ca 


ca 


o 


B 


Q 


OJ 


XI 


■H 


3 


3 


•H 


w 


p 


O 


3 


O 




QJ 






a 


O 


E 


E 


t*s 


X 





QJ 


00 


00 


00 


i—l 


o 


O 


•H 


O 


00 


5-1 




> 


o 




QJ 


s 


CO 


CO 


u 


































QJ 




i—i 


o 


■H 


■H 


w 


ca 


ca 


ca 


CO 


CO 


CO 


CO 


CO 


ca 


• 


• 


• 


• 


• 




CO 


i—l 




X 


1— 1 


ex 


(X 


o 




E 


E 


E 


E 


E 


E 


E 


E 


a 


ex 


(X 


a 


a. 




QJ 


QJ 




o 


:>, 


o 


a 




3 


3 


3 


3 


3 


3 


3 


3 


3 


CO 


CO 


CO 


CO 


CO 


CO 


JJ 


ca 




4J 




1—1 


i— i 


J 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


CO 


X 


X 


X 


X 


X 


Q 


O 


E 


CO 


a, 


00 


o 


o 


< 


1—1 


1—1 


1—1 


i—l 


1—1 


i—l 


1—1 


1—1 


r- 1 


3 


3 


3 


3 


3 


OS 


r— 1 


ca 


:*S 


^ 




•i-H 


■H 


s 


QJ 


aj 


01 


QJ 


.QJ 


QJ 


01 


QJ 


aj 


CO 


CO 


CO 


CO 


CO 


< 


CO 


<-; 


£2 


J-l 


5 


QJ 


QJ 


QS 


X 


X 


p 


_T^ 


-G 


^ 


p 


r-| 


*~ 












SI 


CJ 


O 


l—l 


o 


e> 


—1 


rJ 


5 


04 


O-i 


0-1 


04 


Oj 


a< 


04 


04 


O-i 








-]c 




1— 1 






si 










1— 1 






























.-J 






CO 










a 































c 
















































o 

■H 

c 

3 








^4^4 




M 






M 






* o 










^ 






QJ 
















































CO 
















































CD 
















































3 
















































oo 

■H 








^ 


£ 


^ 






















oo 








-o 
















































O 
















































PC 
















































X 
















































13 3 
















































3 CO 

3 i-l 




* 






























o c 












o w 
















































OS r-l 
















































CO 
















































3 
















































•H 
















































4-1 
•H 

5-1 
-j 




o 


«<*««^ 




< 




's 


<4 






r-. 




o o 






* 




<A 


CO 


















































































































' C 
















































C 










































^~s 






u 
















^— s 


























co 






X 
















c^. 








^-N 


















3 






•r-l 
























CO 


















-H 






CQ 
















CO 








QJ 


















U 






















4-1 








4-1 


















o 






o3 
















CO 








c 


















4-1 






















J-l 








CO 


















CO 






r-l 
















J-l 








oo 


















o 






•r-l 
















QJ 








■H 


















r4 






u 
















CO 








OO 


















QJ 






^QJ 
















•H 


























E 






E 
3 
















J-l 

4-1 








o 

X 


















II 




C 

o 


CO 
















II 




5 




3 

4-1 

CO 










^~s 








^"-v 




u 


•H 




















j-i 




QJ 










QJ 








CO 


S-l 


XI 


CO 
















s-** 




X 




H 










00 








-0) 


<D 


•H 


C 
















c 




•H 














X 








c 


j*: 


CQ 


QJ 
















O 




DQ 




II 










•H 








c 


aj 




•r-l 
















U 


















Sj 








o 


cu 


4-1 


c 
















X 




4-1 














01 








-) 


i—l 


QJ 


QJ 
















•r-l 




QJ 


,*— s 


/ — ^ 










> 










CQ 




5-1 
















CO 






u 


• 




-^ 






o 








QJ 




r-t 


CO 




















r-l 


QJ 


.a 




u 






J 






rH 

cu 

00 


XI 

3 


CO 


•H 

5-1 


O 

CO 

CO 
















4-1 

QJ 


5-1 


•H 

r4 

*QJ 


OO 

c 
•1-1 


°3 




o 

5-i 




^^% 


CO 






<u 


CO 


^ 


E 


E 














S-i 


r-l 


QJ 


E 


N 


• 




QJ S 

X X 




r. 


i— 1 


!-i 


/ — •* 


.—1 


CD 


4-1 


3 














^-v 


3 


■H 


.C 


3 


4-1 


Q 






to 


3 


QJ 


c 


jz 


J-l 




a 


CO 












S 


O 


5-i 


4-) 


Q 


•H 


s^,,^ 




/QJ CJ 




CO 


CO 


OO 


c 


o 


O 


CO 




■H 












CO 


CQ 


*QJ 


c 


■^_x 


b- 






Cl ■•-< 




>, 


C 


j-> 


CO 


co 


s 


3 


CO 


CO 












JZ 




E 


:3 






CO 




QJ CO 




3 


•H 


4-1 


e 




^^ 


a 


•i-i 


C 












CO 


CO 


3 


O 


CO 




C 




O 




-Q 


co 


QJ 


oO 


CO 




>s 


u 


QJ 






5-1 






s^ ^ 


O 


a 




QJ 


•H 


•r-l 


CO 


^ t 




CO 


c 


O 


01 


3 


CO 


4-1 


XI 


•H 




x— v 


QJ 








•H 


>*_• 




4-1 


U 


4-1 


c» 




2: 


cu 


CQ 


■H 


4.) 


•H 




3 


C 




^QJ 


3 






'0 


C 




CO 


CO 


cu 


c 


w 


QJ 




r. 


0- 




s 


CO 


3 


CO 


op 


QJ 




C 


o 






4-J 


r8 


•r-l 


4J 


CO 


CO 


CO 


•^ 


C 




•^^ 


u 


CO 


s — / 


c 


5 


3 




-4 




c 


P-. 






CO 


•rH 


a 


4-1 


E 


JZ 


o 


sj x: 






qj 


•H 




a; 


— H 


1— 1 


CO 




•r4 








•1-1 


l-l 


CO 


QJ 


1—1 


CO 


Cl 


■w 


QJ o 




CO 


en 


3 


CO 


u 


CO 


^ 




a 




rJ 


CO 






X 


-8 


J-i 


C 


QJ 


o 


QJ 


sv, 


CO ctj 




Q 




oO 


4_) 


U-l 


E 


4-1 


Ifi 


CO 




%^ 


•H 




CO 


CO 


00 


•H 


G. 


> 


i — I 


p 


•H T3 




PS 


w 


C 


CO 






o 


3 


CO 






r-l 




ca 


5-i 












QJ 


&H 


3 C 




< 


3 


3 


rH 


co 


en 


CO 


rH 


E 




CO 


CO 




CO 




CO 


CO 


CO 


CO 


CO 






X -H 




N 


i— i 


c 


■r-l 


3 


3 


X 


>^ 






3 


U 




•<^ 


CO 


•H 


•r-l 


•H 


•H 


•i-l 


CO 


rH 


3 QJ 




i— I 


>-. 


■H 


4-1 


■—1 


>— i 


o 


4-1 


CO 




C 


3 




o 


^ 


a 


Q- 


CX 


a 


a 


>1 


<» 


CO 4-1 




J 


j-j 




3 


>s 


>* 


r-l 


a 


C 




•H 


4-1 




+i 


i—l 


CO 


CO 


CO 


CO 


CO 


rH 


+> 


CO 






o 


CO 


E 


4.) 


4-1 


1— 1 


CO 


QJ 




J-l 


4-1 




fc 


QJ 


CO 


CO 


CO 


CO 


CO 


QJ 


C3 


CO 




J 


CO 


■H 




O 


o 


-^ 


X) 


X 




CO 


3 


CO 


o 


JZ 


5-1 


5-4 


u 


5-1 


J-l 


JZ 


.■i 


O X 




< 


X 


> 


CO 


CO 


CO 


X 


o 


CO 




E 


00 


fcj 


Ei 


O 


X 


X 


X 


X 


X 


o 


1 


■H >s 




2 


o 


CO 


S-i 


X 


X 


a 


X 


_C 








CO 


1 


o 


C 


3 


C 


C 


c 


o 


-s; 


CO 3 




as 


XJ 


C 


^ 


•H 


•r-l 


•r-l 


•H 


CJ 




O 


o 


r-l 


*3 


S-i 


■H 


•H 


•H 


•H 


•H 


CO 


CO 


3 O 




33 


SJ 


Q) 


-C 


E 


E 


E 


Q, 


>, 


to 


U-l 


M-l 


O 


ss 


4-1 


r-l 


i— 1 


r-l 


r-l 


i—l 


tx 


cu 


— 1 -H 




G 


^ 


X 


cu 


0) 


0) 


QJ 


QJ 


O 4-1 


Q 


3 


3 


H 


« 


CO 


^ 


^ 


>, 


>, 


:>, 


•H 


fc 


QJ S-l 




CJ 


W 


o 


T" 


"'"I 


X 


J 


O Pa 


< 


CQ 


CQ 


QS 


l-q 


-c£ 


o 


u 


u 


u 


U 


Q 


k 


Cu H 




o 
















g 


En- 






B 


1 
















I 





6 

■l-l 




* 




* 


^ 


4 


c 














3 














\QJ 














Pi 














CO 














CU 














3 
•H 




• 




• 




<4 


n 














XI 














O 














rv^ 














T3 














XI c 
C TO 




* 




* 


* * 




3 --I 














O CO 














C£ 1— 1 














CO 














3 














•H 

4-1 

•H 

3 




* 


* 


* 


O^* 


<4o 














CO 














s 


























cu 














^ 










^^ 




CO 










1—1 




c 










1—1 




CO 










■H 














X 


^^ 


XI 

c 








^^^ 


CO 


CO 


-H 






CD 


QJ 

4-1 


7^ 


5 


i—l 
X 






r— 1 


3 


& 


w a) 


' — ^ 






4-> 


i—l 


rt 


^ 








u 


v ' 




CO 


^*N 






3 




s ' 


i — ii — ii — i c 


c 






4-1 


•H 


^^ 


CO 


•H 
XI 






c 


r— 1 


VQJ 


•-n CO 


3 U 






a) 


i— ( 




0) ^ 0) 


co cu 






0) 


<D 


c 


•H i—l co 


Q J- 1 






M 


XI 


•H 


O CJ w 


^^ 4-1 






00 


CO 

> 


XI 
CO 


co o v ai a 


nus 
fste 






*CD 




4-1 


4-> CO c •■- 


•H M-l 






C 


CO 


CO 


cfl v_^ C -i-l 


E O 






c 


CU 


a 


C -H -J 


CO 33 






•H 


O 


•H 


■H -H XI v-x 


y 






XI 


CO 


u 


H V4 w 


X -H 








•H 


X> 


co a) co 


QJ 








u 


E 


O -H CO 3 


CO -H 




CO 


CO 


o 


• H 


O E 3 H 


a n 




Cfl 


CO 


o 




4-1 3 O 3 


O CO 




r-i 


XI 




CO 


■—I CO -H 4-1 


.-I o 




+i 


^ 


CO 


>, 


3 co i— I cO 


X 




Sh 


E 


^ 


i— 1 


E 3 3 rH 


a : 




3 




i— 1 


CU 


X) cfl Oj 


>, CO 




Eh 


CO 


CU 


x 


CO 


4-1 Ch 






■H 


X 


o 


•i-l CO C co 


o o 




<» 


q 


O 


o 


J-l CD O -H 


X i-l 




e 


o 


O 


E 


01 U XI E 


Oj X 




•ri 


i— i 


E 


4-> 


CO >, CO O co 


s a. 




Sh 


0) 


H 


a) 


W i—l CO O ■— 1 


cO >^ 




a 


x 


CU 


K 


:*: o eg >* cu 
< aq CJ X cl. 


£ = H 




% 


u 


a 


W 












3 






i 








co 





CO 

cu 

•H 

o 
cu 
a, 

CO 

o 

■H 
E 
CU 

XI 



c 
cu 

CO 

cu 

M 



XI 

cu 

N 
•H 
i-H 
CO 
cJ 
O 



CO 

4-1 

c 
a) 

T3 

•1-1 
o 
a 

cfl 



* o 



CO 
01 
•H 
CJ 

cu 

CO 

XI 
0) 

o 

3 
XI 

o 

u 

4-1 

c 



c 
cu 

CO 

cu 

M 
CX 



XI 
0) 
N 
•H 
i—l 
CO 

o 

o 



^ * < 



2 



54 



AMPHIBIANS AND REPTILES ERRONEOUSLY CITED 
OR ACCIDENTLY ENCOUNTERED IN THE MASCARENE ISLANDS 



M = Mauritius 



Reunion 



Rodr. = Rodrigues 



FROGS AND TOADS 

- Bufo melanostictus 

- Rana hexadactyla 

- Rana tigerina 



CROCODILES 

- Crocodylus niloticus 

- Crocodylus porosus 



TURTLES 

- Amyda cartilaginea M 

- Chinemys reevesii M 

- Chrysemys s. elegans M, R 

- Cuora amboinensis M 

- Dipsochelys elephant ina Rodr. 

- Geomyda spengleri M, R 

- Homopus areolatus M 

- Homopus signatus M 

- Psammobates geometricus M 

- Psammobates tentorius M 

- Testudo graeca Rodr. 



SNAKES 

- Acrantophis dumerilii M 

- Dromicodryas bernieri M 

- Helmintophis flavoterminatus M 

- Liophidium vaillantii M 

- Praepeditus lineatus M 



CHAMELEONS 








- Chamaeleo 


bif idus 


M, 


R 


- Chamaeleo 


lateralis 


R 




- Chamaeleo 


pardalis 


M 




- Chamaeleo 


parsonii 


M 




- Chamaeleo 


pumilus 


M, 


R 


- Chamaeleo 


tigris 


M, 


R 


- Chamaeleo 


verrucosus 


M, 


R 



(an obscure name; the identity of 
this nominal species has not yet 
been clarified) 



55 



REFERENCES 



Arnold, E.N. (1979). Indian Ocean giant tortoises : their systematics 

and island adaptations. Phil. Trans. R. Soc. Lond. (B) 286 : 127-145. 

Arnold, E.N. (1980). Recently extinct reptile populations from Mauritius 
and Reunion, Indian Ocean. J. Zool., Lond. 191 : 33-47. 

Bour, R. (1978). Les tortues des Mascareignes : description d'une espece 
nouvelle d'apres un document (Memoires de l'Academie) de 1737 dans 
lequel le crane est figure. C.R. Acad. Sc. Paris (D) 287 : 491-493. 

Bour, R. (1979). Premiere decouverte de restes osseux de la tortus 

terrestre de La Reunion, Cylindraspis borbonica. C.R. Acad. Sc. Paris 
(D) 228 : 1223-1226. 

Bour, R. (1980). Systematique des tortues terrestres des iles Mascareignes : 
genre Cylindraspis Fitzinger, 1835 (Reptilia, Chelonii). Bull. Mus. 
natn. Hist. nat. Paris, 4e sec, A, 3 : 895-904. 

Bour, R. (1980). Essai sur la taxinomie des Testudinidae actuels (Reptilia, 
Chelonii). Bull. Mus. natn. Hist, nat., Paris, 4e ser., A, 2 : 541-546. 

Bour, R. & F. Moutou (1982). Reptiles et amphibiens de l'ile de la Reunion. 
Info-Nature, He de la Reunion, n° 19 : 119-156. 

Vinson, J. & Vinson, J.M. (1969). The saurian fauna of the Mascarene Islands. 
Bull. Mauritius Inst. 6 : 203-320. 




c 

CO 
0) 

o 
o 

c 

CO 

•H 

T3 
C 






CO 

c 

CO 



C 

ai 

CO 

o 

CO 

o 

si 



•H 



57 30 
— Subfossil carapaces,! 
* bones of reptiles collected. 

I 
q Living specimens reported by 
old settlers, sailors etc. 




MAURITIUS 




RODRIGUES 



MARINE AND TERRESTRIAL FLORA AND FAUNA NOTES 
ON SOMBRERO ISLAND IN THE CARIBBEAN 

by Nancy B. Ogden, William G. Gladfelter 

John C. Ogden and Elizabeth H. Gladfelter* 



Introduction 

Sombrero Island (Lat. 18° 36' N, Long. 63° 25' W; Fig. l) is an 
elevated block of limestone 6-12 m above sea level with no beaches and 
a precarious anchorage subject to damaging ground swells. It is 1.2 km 
long and approaches the shape of an obtuse triangle extending northeast 
and southwest. Isolated from the Anguillan bank by the 32-mile Sombrero 
Passage, it forms the northernmost limit of the Lesser Antilles and is 
separated from the Puerto Rican/Virgin Islands bank (part of the Greater 
Antilles) by the 40-mile wide Anegada Passage. Sombrero is 366 m wide 
at its widest and represents a remnant of an island believed to have 
been as large as its present 5.6 x 8 km wide underwater platform which 
varies from 16-30 m deep. It is believed to consist of a volcanic base 
capped by Pleistocene limestone (Julien, 1866). According to Julien 
this limestone represents the floor of a pre-historic lagoon once 
protected by a barrier reef and possibly enclosed by an atoll. His 
evidence for this is the abundance of fossils, in particular, Bulla 
(bubble shell). The fossil shells and corals are extensively described 
by Julien. 

Presently Sombrero is exposed to the open Atlantic Ocean. The 
occasional severe ground swells and regular heavy seas limit coral 
growth. They also affect the land-dwelling populations. Fortunately, 
the seas were calm for our stay, but even then the salt spray was 
evident. The eroded cliffs are precipitous (Fig. 2) and undercut, but 
occasionally slope toward the water and are then undercut below the 
water surface. The island was uninhabited until 1856 when operations 
began for mining of the phosphate rich, rock-guano deposits. Presently 
it is inhabited by four lighthouse keepers. 

The following observations were made on June 10-11, 1979 using the 
yacht Tiichtig . 



* West Indies Laboratory, Fairleigh Dickinson University, Teague Bay, 
Christiansted, St. Croix, U. S. Virgin Islands 00820. 

Atoll Res. Bull. No. 292: 61-74, 1985. 





SOMBRERO 


ISLAND 


22 


16 
20 

/ 

/ M 

s ^ f 

TUCHTIG X 0^' r^ 
ANCHORAGE ''1/^* R 

/ wa 1 
/ f\ 1 

/ I V 

'/ Old quames 


,-^) «. POINT WOOD 
/•' /U^O 29 

/I 1 

" I 25 


-18°36' 




v (6 -12m elev.) f 


- 




Landing / 
22 V 


LIGHT ^-^0 
Ruins * f 

' ^-"-^ Point Ray 


29 

2; 




S> D _7 " 






Ce P 

24 ^oih 








T WARNER 


meters 500 






63J25' 


1 


26 


LEEWAI 


3D ISLANDS 

ANEGAOA 

, SOMBRERO 

> < ^ <= * *^*ANGUILLA 

<ST.MARTIN 

• 


100 km 


\ PUERTO RICO 3 






ST.CROIX " ^ 

CARIBBEAN SEA ° , 0> 








ft 



Fig. 1. Sombrero Is. (DMA, 1976) with an inset of its location in the Eastern Caribbean 

island chain; depths in meters; dotted line shows extent of the underwater survey. 
A-E represents bird areas listed in Table 2. 



63 



land Plants 



One's first impression is that there are only two species of land 
plants, but walking the length of the island and back revealed 8 species 
(Table l). Pluchea symphyt i folium (grows to 3 m tall) must be a recent 
introduction as previous literature refers to the cactus Opuntia 
ant i liana (grows to 0.6 m tall) as the tallest plant on the island. The 
most abundant plant, sea purslane ( Sesuvium portulacastrum ) , appeared 
to be the preferred nesting site for boobies (Fig. 4 ) while blast holes, 
rock rubble, crevices and natural depressions provided numerous niches 
for the other six species of nesting sea birds. 

Birds 

Seabirds arrive from the south in March and depart in late May, 
June, July and even August (Lawrence, 1864). "The lighthouse crew" 
(during phosphate mining years) "reports that nests and eggs can be 
found on the island during at least 8 months of the year" (Lawrence, 
1864). According to Lawrence, the nesting seabirds suffered consider- 
ably from poaching by the phosphate miners. Quarrying of the guano 
deposits went on from 1856 (Lawrence, 1864) to 1890. Julien arrived 
in i860 and four years later made the statement that ".. laborers have 
been so indefatigable that I do not believe a single young bird has 
been hatched since our occupation of the Key. Consequently, last season 
(spring of 1863) only about two dozen eggs in all were found, instead of 
the thousands of previous years." Julien also mentioned wild cats as 
predators (Lawrence, 1864). We suspect the nesting bird populations 
have recovered considerably since then, but according to the present head 
lighthouse keeper, men still come from Anguilla to get eggs and birds. 
Migrating birds sometimes appear on the island and occasionally birds 
from larger neighboring islands (Lawrence, 1864). We observed nests of 
noddy, sooty and bridled terns and brown boobies. Julien never saw 
booby nests despite their presence from June to November - suggesting 
the thorough poaching of their eggs. Lazell (1964) did see them nesting. 
Noddy terns were most abundant of all, as was the case during Julien 's 
stay (Table 2). 

Reptiles 

Other than the seven species of nesting birds, the black lizard 
Amieva corvina (Teiidae) is the next most obvious inhabitant. This 
endemic, slatey black or black and tan (occasionally with light speckled 
flanks) lizard was the only recorded living reptile until 1964. We 
observed it throughout the island. Its main food is reported to be 
bird's eggs (Lazell, 1964). One was observed during our stay eating the 
small, yellow Portulaca flowers. Lazell (1964) reported the lizard 
Anolis gingivinus ( Iguanidae ) and the gecko Sphaerodactylus sputator 
(Gekkonidae) - both are common to the Anguillan and St. Kitts banks. 
Only one Anolis was seen in the north central portion of the island 
during our visit. According to Lazell (1964) the Anolis varies from 
olive drab to distinctly light greenish with a bold, light flank stripe 
and varying spots. The gecko is "at or near the longitudinally-striped, 
extreme color pattern" (Lazell, 1964). It was not seen by us. 



64 

Julien (1866) mentions the discovery of fossil remains of land 
turtles which he was led to believe belonged to three new extinct and 
gigantic species similar to those of the Galapagos Islands. Auffenberg 
(1967) finds that none of Julien 's original fragments can be found, but 
he and a colleague obtained more in the Pleistocene fissures at the 
northern end of Sombrero in 1964. Auffenberg (1967) states that Julien 's 
tortoise fragments were originally described as the new species Emys 
sombrerensis Leidy. It is now known as Geochelone sombrerensis (Leidy) 
Auffenberg, 1967. The sternum length was hypothesized to have been 12 
inches. Some of Julien 's material reportedly represented specimens with 
a plastral length of at least 32 inches (Auff. 1967). Geochelone 
carbonaria (Spix) Morocoy is a species presently found on many Carib- 
bean Islands. It is commonly around 12 inches in length, but larger 
ones have been reported (pers. com.). _G. denticulata (Linneaus) which 
has reached a shell length of 26.5 inches is reportedly found in Trinidad 
(Underwood, 1962). 

Marine Algae 

The marine algae on the exposed limestone cliffs were quite differ- 
ent from other Caribbean islands. Wrangelia penicillata , commonly found 
2 m deep and deeper, was the most common macroalga in the splash zone 
for 2 m above the surface. Also there was a surprising lack of 
Sargassum and Turbinaria at the surface. Protected areas were more 
typical of combinations found elsewhere. In general brown algae - 
Dictyopteris delicatula ; Dictyota dichotoma , JJ. dentata and Lobophora 
variegata represented the most biomass 2 m deep and below. Four species 
of colorful, fleshy, encrusting algae were very striking in the top 
two meters . Snorkelling and scuba diving were limited to the northern 
half of the leeward side (Fig. l), while collections were made from 
land on the windward side of the northeast end (Table 3). 

Coral 

There were no reef-like accumulations of coral, but the species 
common on other eastern Caribbean islands were generally well represen- 
ted particularly on the shelf on the northern side of the island. 
Colonies were generally of moderate size and flattened, especially on 
the vertical slopes of the island. Coral growth must be affected by 
the occasional severe ground swells that toss boulders and cobbles 
about on the bottom and have cut a notch at the base of the island at a 
depth of about 10 m. 

Fishes 

The fish community generally was rich and diverse (Table h). A few 
species unusually well represented were Kyphosus sectatrix (Bermuda cnub) 
and Cephalopholis fulva (coney) - common near the walls; and Paranthias 
furcifer (Creole fish) and Xanthichthys ringeas (Sargassum triggerfish) 
common over the leeward shelf. Conspicuous by their absence were grunts, 
probably due to the lack of soft bottom and seagrass beds. The propor- 
tional representation of major trophic categories (Table 5) was similar 



65 

to shelf edge fish communities censused elsewhere in the northeastern 
Caribbean by the authors (unpublished). 

Miscellaneous Observations 

A colorful variation of the crab Grapsus can be seen skirting the 
edges of the island. The periphery of the island is dotted with salt- 
water, brackish and fresh water pools. Seven meters above the sea 
level there were pools with occasional ocean surgeonfish, sergeant majors 
and Diadema (the long-spined black sea urchin). Echinometra , a short- 
spined urchin, inhabited the wave-washed areas. Some pools were choked 
with filamentous, green algae. Others had water bugs, backswimmers , and 
other small swimming organisms. 

Insects, in general, were not noted, but houseflies, a flattened, 
tan spider, the cabbage butterfly, red mites and small elongate ants 
were observed. Tectarius shells seemed poor fare for the purple-clawed 
hermit crab Coenobita c lypeatus which grows much larger on other islands 
where larger shells are available. In general, other than sea birds, 
terrestrial life is very limited - probably due to the lack of soil and 
the presence of abundant salt spray. 

Acknowledgements 

Thanks are due to Dr. Michael Wynne of the University of Michigan 
for help with a few of the algae. Dr. Edward Towle of Island Resources 
Foundation generously provided us with the reprints and a map of 
Sombrero Island. Dr. F. Raymond Fosberg of Smithsonian looked over 
the land plant collection and encouraged this report. This is 
Contribution No. 12k of the West Indies Laboratory. 



66 



Bibliography 

Auffenberg, W. S. 1967- Notes on West Indian Tortoises. Herpetologica 
23: 3k-kk. 

Defense Mapping Agency 1976. Anguilla to St. Christopher. Hydro- 
graphic Center, Wash. D.C. 20390. Chart No. 25606. 

Dickson, Hanmer (Ms.). Sombrero Island (history acquired from the 
lighthouse keeper). 

Julien, A. A. 1866. XXXI. On the Geology of the Key of Sombrero, 
W. I. Ann. Lyceum Natural History 8: 251-278. 

Lawrence, G. N. 1861*. VIII. Catalogue of Birds Collected at the 
Island of Sombrero, W.I., with Observations by A. A. Julien. 
Ann. Lyceum Natural History 8: 92-106. 

Lazell, J. D. Jr. 196U. The Reptiles of Sombrero, West Indies. 
Copeia k: 716-718. 

Underwood, Garth 1962. Reptiles of the Eastern Caribbean. Dept. 

of Extra-Mural Studies, University of the West Indies, Trinidad. 
Carib. Affairs No. 1: 192 pp. 




Fig. 2. Looking north at the lighthouse near 
Pt. Ray; windward side on a calm day showing 
precipitous nature of the island's shoreline. 




Fig. 3. Quarry area near our anchorage 
looking south toward the lighthouse. 




Fig. h. Area A (Fig. l) with nesting 
brown boobies on the predominant island 
vegetation ( Sesuvium portulacastrum ) ; 
bridled tern also present. 



68 



TABLE I 

LAND PLANTS 
Listed in order of apparent abundance 



Sesuvium portulacastrum (L. ) L. , Aizoaceae - trailing succulent with 
light pink flowers; the most common plant on the island and 
possibly eaten by Amieva . 

Euphorbia mesembrianthemi folia (Chamasyce buxifolia ) Jacq., 

Euphorbiaceae - bushy, yellow-stemmed (basally woody) plant to 
50 cm tall with small, ovate, vaguely succulent leaves in two 
ranks (the leaves strikingly angled); the second most common 
plant on the island. 

Heliotropium curassavicum L. , Boraginaceae - bushy, trailing plant 

with somewhat succulent, frosted looking leaves with small white 
flowers in a typical scorpioid spike; branches grow to 15-1+5 cm 
long; fairly common in moist, protected areas on the southern 
end. 

Portulaca oleracea L. , Portulacaceae - low, somewhat succulent herb 
found in an isolated patch; Amieva was observed eating the small 
yellow flowers . 

Euphorbia serpens H.B.K., Euphorbiaceae - fragile, pink-stemmed, 
trailing herb with small ovate leaves; found with Portulaca . 

Opuntia ant i liana Britton & Rose, Cactaceae - yellow-flowering cactus 
with 3-6 coarse, yellow spines; grows to 0.6 m tall; found iso- 
lated and locally abundant in a large, sloping pit. 

Fimbristylis cymosa Roth, Cyperaceae - a low sedge somewhat 2-ranked 
at the base; blades slightly cupped basally; spikelets 3 mm long; 
found between the lighthouse and the landing. 

Pluchea symphyti folium (Mill) Gillis, Compositae - a bush to 3 m tall 
with large, lanceolate leaves that smell like horse liniment; two 
bushes seen in the sandy area N.E. of the buildings; probably a 
recent introduction. 



69 

TABLE 2 

BIRDS: Comparative observations of 
Julien (Lawrence 18610 appear in parentheses 

Refer to Fig. 1 for lettered areas 

Noddy tern ( Anous stolidus ) - nesting throughout in caves and crevices; 
100 birds seen on nests; one egg per nest; probably 500 nests on 
the island ( Julien describes the nest-building process); birds 
are aggressive. 

Sooty tern (Sterna fuscata ) - some nesting, mostly not (?); concentrated 
in area north of the lighthouse; about 100 birds total (associate 
with noddies ) . Area B_. 

Bridled tern (S . anaethetus ) - mostly south of the lighthouse; some 
nesting; more common than sooty terns; probably several hundred 
birds total; many dead and sick among the rocks (Julien suggests 
it to have been the most abundant bird next to the noddies and 
royal terns; lays one* egg). * We saw up to three eggs per nest. 
Area _D. 

Roseate tern (S_. dougallii ) - one main colony in the south - a "grassy" 
area; not nesting, but probably do; a few were observed roosting 
on tailings mid-island; about 30-i+0 total. Area E_. 

Least tern (S . albifrons ) - main concentration in the center of the 

island in a flat, pebbly area; no sign of breeding, but probably 
do; 20-30 birds total. Area _C. 

Royal tern (Thalasseus maximus ) - one pair observed arriving in the 
evening and departing in the morning (nested in June; departed 
in August; present in variable numbers; disappearing during the 
day or for weeks ) . 

Tropicbird ( Phaeton sp. ) - one seen overhead in the evening. (Julien 

saw one nest and small numbers of P. aether eus , but suggests they 
were abundant previous to mining years. He also described seeing 
one other species). 

Laughing gull (Larus atricilla ) - common on leeward bluffs (not abundant 
during Julien' s stay; associating with the royal terns.) 

Frigatebird (Fregata magnificens ) - common (breed June - July). 

Brown Booby (Sula leucogaster ) - nesting at northern tip of the island; 
eggs and large young present; about 30 nests (?); roost on leeward 
rocks and off south end of island; about 60 birds total. Area A. 
(Fig. h). 

Ruddy turnstone (Arenaria interpres ) - 1 pair seen at several sites 

on the island (small numbers present November - May). 
Cattle egret ( Bubulcus ibis ) - T in the vicinity of the lighthouse. 



70 



TABLE 3 

MARINE ALGAE 



Rhodophyta (red algae) 



Amphiroa sp. (NllUM 0.5 m deep, small inlet 

Botryocladia sp. 16 m deep 

Callithamnion herveyi (N11U2, N11TT) uncommon, 1.5 m deep in small 

inlet and 5 m deep on wall. 
Centroceras clavTilatum (N1185) splash zone 

Ceramium byssoideum (on N11^3, NII69 & N1185 ) splash zone to 1.5 m deep. 
Champia parvula ( on N1151) 13 m deep 
Chondria sp. ["erect on N1151) 13 m deep 
Chondria sp. (creeping on N1151) 13 m deep 
Corallina subulata (N115M lh m deep 
Crustose coralline cobbles (unidentified) 16 m deep 
Gelidium pus ilium 16 m deep 
Griff ithsia sp. (on N1151) 13 m deep 

Grallatoria reptans (N1173) abundant exposed and to 1 m deep 
Halymenia pseudof lores ia (N1146, N11T0, N1153) 5~l6 m. had a striking 

bright orange on brown reticulation which turned plain rose-colored 

upon preservation. 
Herposiphonia pecten-veneris (N11T6) 16 m deep 
Hypnea sp. l6 m deep 

Hypoglossum tenui folium V. carolinianum (on N1151 & N115M 13 m deep 
Jania adherans (with Amphiroa NllUU) 6T5 m deep 
Jania capillacea (on Laurencia obtusa N11U3) 1 m deep 
Laurencia intricata (Nll60, N1185) 1 m & less in inlet and splash zone. 
Laurencia obtusa (NILU3, NII69) scattered to abundant - surface to 1.5 m 

deep 
Laurencia papillosa (N1183) exposed in splash zone 
Liagora valida (N1180) 1^ m deep 
Lophosiphonia sp. splash zone 
Martens ia pavonia (on N1151) 13 m deep, also seen protected near the 

surface. 
Polys iphonia sp. (N1176) from Colpomenia 16 m deep 
Taenioma pur pus ilium (N1185) splash zone 
Wrangelia argus seen at the surface 
Wrangelia pencillata (N11T1, N1185 ) 1-2 m above the surface dominating a 

1 m wide band in the splash zone 
3 unidentified fleshy crusts - bright orange (most abundant), rose-pink 

and chocolate purple; very dominant from the surface to 2 m deep. 
Unidentified polysiphonous red - 16 m deep, creeping on Lobophor a (N11T3) 

Fhaeophyta (brown algae) 

Colpomenia sinuosa (N11T9) 16 m deep, small plants 

Dictyopteris delicatula (NII65, N11U8) the most dominant alga from 2-3 m 

deep, but also found at 1^ m deep. 
Dictyota bartayreesii (N1168) I** m deep 
D. cili olata (N1182) lU m deep 



71 
Phaeophyta (brown algae) continued 

I), dentata (N1167) scattered to abundant several meters deep; also 

abundant 9 - 1^ m deep. 
D. dichotoma (Nll^T) very abundant 9-1G m deep 
D. divaricata (N1182) 16 m deep 
D. sp. (N1164) 16 m deep 

Dllophus alternans (Nll66) 16 m deep, stunted 
Lobophora varlegata (with N115M (N1163) abundant 16 m deep; common as 

a crust in shallow water and exposed. 
Padina sp. (N11T8) 16 m deep, scarce 
Sargassum polyceratium (NllUl) scattered or locally abundant, 1 m deep 

to exposed. 
Sargassum sp. (Nll62) 9 m deep on wall 
Sphacelaria sp. (N1176) 16 m deep 
Stypo podium zonale (Nll6l) small pint 16 m deep 
Turbinaria turbinata (NllUO) scattered, surface to exposed. 

Chloropnyta (green algae) 

Acetabularia sp. l6 m deep 

Anadyomene stellata (NIIU9) abundant on parts of the wall 5 - 8 m deep 

Avrainvillea nigricans (N1150) 13 m deep 

Bryopsis plumosa abundant in one inlet at the surface 

Caulerpa microphysa (N1155) 1.5 m and \h m deep 

Caulerpa vickersiae l6 m deep 

Cladophora sp. in tide pool 

Halimeda discoidea (N1152) 13 m deep 

Halimeda tuna (N1151) 13 m deep 

Neomeris annulata scattered l6 m deep 

Rhipiliopsis stri (N1176) on wall 8 m deep 

N1174 unidentified long skeins of very fine green filaments (l7u diam) ; 

spray zone tide pool. 
N1175 unidentified, branched, tangled, filamentous green 8 cm tall; 

spray zone tide pool. 

Cyanopnyta (bluegreen algae) 

Trichodesmium thiebautii (N1186) quite common; planktonic 

N1181 

N1184 

Oscillatoria sp. (N1176) common on wall 



72 



TABLE h 

FISHES 

Feeding and Abundance Categories (Sombrero Island, 
Northern Half of Leeward Side) 



Feeding Category Key: 



Abundance Category Key: 



C = Crustaceavore 

P = Planktivore 

F = Piscivore 

H = Herbivore 

G = Invertebrate Generalist 

S = Invertebrate Specialist 



1 


= 0-1/hr 


2 


= 1-2 hr 


3 


= 2-5 


k 


= 5-10 


5 


= 10-25 


6 


= 25-30 


7 


= 50 



C Holocentrus rufus (squirrelfish) - mainly in boulder 

strewn cove. 
C 5 H. ascensionis (long jaw squirrelfish) - base of wall; more 

common toward the seaward end. 
C 3 Ad i oryx vexillarius (dusky squirrelfish) - among boulders 

in cove. 
CI A. coruscus (reef squirrelfish) - base of leeward wall. 
C Flammeo mar i anus (longspine squirrelfish) - base of cliffs. 

P h Myripristis jacobus (blackbar soldier fish) - boulders in 

cove. 
P 3 Priacanthus cruentatus (glass eye snapper) - boulders in 

cove. 
F 5 Lutjanus apodus (schoolmaster) - in schools on reef flats 

and near cliff base at seaward end. 
F 2 Mycteroperca venenosa (yellowfin grouper) - large, near 

base of wall and out on reef flats. 
F 2 M. tigris (tiger grouper) - large, at cliff base. 
F 1 Epinephelus adscensionis (rock hind) - boulders in cove. 
F 1 E_. striatus ( Nas s au grouper ) - under ledge near second 

mooring. 
F 6 Cephalopholis fulva (coney) - along walls, base of walls 

and rubble. 
P 6 Paranthias furcifer ( Creole fish) - over reef flats and 

south end of wall. 
P 5 Gramma loreto (fairy basslet) - under ledges 
C 2 S err anus tigrinus (harlequin bass) - rubble at base of 

cliffs. 
F 2 Caranx latus (horse-eye jack) - over sand, south end of 

swim; chasing mackerel scads. 
F 5 _C. ruber (bar jack) - in water column near moorings. 
P 6 Decapterus macarellus (mackerel scad) - in a few schools 

above reef flats. 
F 2 Seriola rivoliana (almaco jack) - same as horse-eye jack. 
F k Sphyraena barracuda ( great barracuda ) - along walls and 

over reef flats. 



73 
FISHES (continued) 

H 5 - 6 Kyphosus sectatrix (Bermuda chub) - cliffs, base of cliffs 

and over reef flats . 
Hemiramphus balao (balao) - near surface at first mooring. 
Gymnothorax sp. (moray) 
Malacanthus plumieri (sand tilefish) - sand and rubble 

areas . 
Mulloidichthys martinicus (yellow goat fish) - one group 

in boulders of cove. 
Pseudupeneus maculatus (spotted goatfish) - small junction 

of reef flats and sand. 
Calamus sp. (porgy) - over sand flats 
Amblycirrhitus pinos (redspotted hawkfish) - sides and 

base of wall. 
Equetus punctatus (spotted drum) 
Holacanthus ciliaris (queen angelfish) - cliff base and 

walls at seaward end. 
H_. tricolor (rock beauty) - cliff base and outer (seaward) 

reef flats. 
Pomacanthus arcuatus (grey angelfish) - one pair in 

shallow sand-bottom cove. 
_P. paru (French angelfish) - reef flats and cliff base at 

seaward end. 
Eupomac entrus dorsopunicans (dusky damselfish) - juveniles 

common on walls. 
EL partitas (bicolor damselfish) - common on deeper reef 

flats near seaward end. 
Microspathodon chrysurus (yellowtail damselfish) - boulders 

in cove. 
Abudefduf saxatilis (sergeant major) - inshore. 
Chaetodon striatus (banded butter fly fish) - cliff base and 

walls at seaward end. 
P 5-6 Chromis cyanea (blue chromis ) - reef flats and cliffs at 

seaward end. 
P 5-6 _C. mult ili neat a (brown chromis) - cliffs 
P,G 6-7 Thalassoma bifasciatum (bluehead) - common 

G 2 Halichoeres bivittatus (slippery dick) - sand and reef flats. 
G h H_. gar not i (yellowhead wrasse) - cliff base, especially the 

seaward end. 
G 3 II. maculi pinna (clown wrasse) - cliff base. 
G 2 H. radiatus (puddingwife) - cliff base 
P 5 Clepticus parrai (creole wrasse) - water column above reef 

flats. 
G 5 Bodianus rufus (Spanish hogfish) - small to large along 

cliffs and cliff bases. 
Scarus coelestinus (midnight parrotfish) - reef flats. 
S croicensis (striped parrotfish) - cliff base especially 

the seaward end. 
_S. vetula (queen parrotfish) - cliff sides and bases 
Spar i soma aurofrenatum (redband parrotfish) - cliff sides 

and bas es . 



p 


4-5 


F 


1 


G 


h 


G 


k 


G 


3 


G 


1 


C 


3 


C 


1 


S 


2 


S 


3 


S 


2 


S 


2 


H 


5 


G 


5 


H 


2 


G,P 


5-6 


G 


2 



H 


2 


H 


k 


H 


h-5 


H 


2 



74 



FISHES continued, 



H 


3 


H 


1 


H 


4-5 


H 


5 


H 


3 


H 


4-5 


H 


4 


H 


6-7 


S 


3 


P? 


5 



4-5 



S. chrysopterum (redtail parrotfish) - cliff base. 

_S. rubripinne (yellowtail parrotfish) - reef base. 

_S viride (stoplight parrotfish) - cliff sides and bases. 

Ophioblennius atlanticus (redlip blenny) - cliff sides 

Bathygobius sp~ (goby) - splash zones on the north end. 

Acanthurus bahianus (ocean surgeon) 

A. chirurgus (doctorfish) - seaward reef flats 

A. coerulens (bluetang) - cliffs. 

Balistes vetula (queen trigger fish) - over reef flats. 

Xanthichthys ringens (Sargassum triggerfish) - over reef 

flats and in water column. 
Melichthys niger (black durgon) - over reef flats at 

moorings. 
Cantherhines macrocerus (whitespottted filefish) - large, 

at cliff base. 
_C. pullus (orangespotted filefish) - cliffs. 
Alutera scripta (scrawled filefish) - cliffs. 
Lactophrys bicaudalis (spotted trunkfish) - cliff base. 
L. triqueter (smooth trunkfish) - cliff base 
Diodon hystrix (porcupinefish) - cliff base in a cave. 



t artj: 5 

Summary of Abundances of Fish Feeding Guilds 







No. spp. 


% of Total 


Category 




Feeding 




per 


No. of 


Abun- 


% of Total 


Category 


Symbol 


Category 


Species 


dances 


Fish Seen 


Herbivores 


H 


16 


24 


59 


26 


Invertebrate 
generalists 


G 


18 


26 


51 


23 


Invertebrate 
specialists 


S 


5 


7 


12 


5 


Crus tac eavor es 


C 


8 


12 


21 


9 


Piscivores 


F 


11 


16 


31 


Ik 


Planktivores 


P 


10 


15 


50 


22 


Total 




68 




224 





VEGETATION AND FLORA OF THE LOWENDAL ISLANDS, 
WESTERN AUSTRALIA 

by Ralf Buckley 



The Lowendal Islands are a group of small limestone islands between 
Barrow Island and the adjacent mainland of north western Western 
Australia (Figure 1). They lie between 20° 34' 30" S and 20°41'11"S and 
between 115°30'17"E and 115°34'43"E. Since the islands do not have 
individual names, they are designated by letters in Fig.1. They 
comprise eroded remnants of Pleistocene reefs, with caps, dunes or 
beaches of Holocene calcareous sand in some cases. In area, they 
range from bare pinnacles of a few hundred square metres or less to a 
relatively complex island of 0.77 km 2 . This is named island B in 
Figure 1: no particular significance attaches to the order of naming. 
There are two main limestone terraces, the lower presumably of 
Holocene age: their precise elevations above mean sea level have not 
been ascertained. The climate corresponds to that of Barrow Island 
(Buckley 1983): a monsoonal climate with mean annual rainfall 200 mm, 
peaking in February-March and May-June. The only previous botanical 
record for the Lowendal islands appears to be that of Serventy and 
Marshall (1964), who recorded five species, namely Ptilotus exaltatus , 
Scaevola spinescens , Myoporum acuminatum , Spinifex lonqifolius and 
Triodia sp. from "Lowendall Island", presumably island B of Fig.1, in 
September 1958. 



According to Serventy and Marshall (1964), "Lowendall Island" was 
named by the French Baudin expedition in March 1803. The official 
account of the expedition made no mention of the island's natural 
history, and it appears that Serventy and Marshall were in fact the 
first naturalists to visit the main island, and Mr W.H. Butler and 
myself the first to visit the smaller islands. The British atomic 
weapons expedition in 1952 visited the island, but it is not included 
in the resulting natural history report by Hill (1955). Serventy and 
Marshall (1964) described the main island as follows: "The island has 
a more picturesque appearance than Barrow Island, with a higher and 
steeper coastline and an attractive beach on the east side which 
affords an easy landing. Much of the island is of "pipey" bare 
limestone but there are vegetated depressions of deep red-brown soil. 
Dunes of white sand have built up at the north end. The flora of the 
island has not the barren aspect that characterises Barrow Island". 
The main aim of their visit was to record the fauna. They found two 
reptile species, namely Phvsiqnathus qilberti and Varanus sp.; no 
mammals, despite a nocturnal search for likely species; and 16 bird 
species, listed in Table 1. Only the first and last of the bird 
species listed were recorded as nesting. 

ASPECT Consultant Group, P.O. Box 114, Eastwood, SA 5063, Australia. 
Atoll Res. Bull. 292: 75-82, 1985. 



76 



The vegetation of the nearby and much larger Barrow Island (233 km 2 in 
area), was mapped and described by Buckley (1983) in terms of 7 main 
vegetation types and 29 subtypes. Five of these subtypes occur on the 
Lowendal Islands, as follows. 



Type 4: Coastal sand assemblages 

Subtype 4a: strandline assemblage; characterised by Ipomoea 
pescaprae and Salsola kali . 

Subtype 4b: white aeolian foredune areas: open vegetation 
dominated by Spinifex longjfolius and also 
characterised by Ptilotus villosif lorus and 
Cynanchum f loribundum . 

Type 5: Coastal rock assemblages 

Subtype 5a: limestone or conglomerate cliffs, scarps and 
headlands bearing Triodia wiseana, and also 
characterised by Sarcostemma australe and Capparis 
spinosa . 

Subtype 5b: low limestone areas bearing Frankenia . Sclerolaena . 
Neobassia and Halosarcia species. 



Type 6: Mangroves 
Subtype 6b 



old stands of Avicennia marina swamped by sand, 
with ground cover of halophytes toward inland 
margin: Neobassia astrocarpa , Sclerolaena spinosa , 
Halosarcia spp., Frankenia paucif lora , Threlkeldia 
diffusa , Enchylaena tomentosa and Sporobolus 
virqinicus . 



On the Lowendal Islands, the primary substrate and vegetation division 
is between limestone and sand. Limestone areas, which correspond 
approximately to the coastal rock assemblages on Barrow Island 
(vegetation types 5a and 5e of Buckley 1983) may be subdivided into 
low terrace, high terrace cliff-tops, and in the larger islands, high 
terrace areas further from the sea. The vegetation of white sands 
corresponds approximately to the coastal sand assemblages on Barrow 
Island (vegetation types 4a, 4b), and may be divided into beaches, 
small dunes, and thin caps over limestone. There are also stands of 
the mangrove Avicennia marina (type 6b) on several islands. 



77 

The plants present on 12 of the islands were recorded during visits in 
September 1980; these records are summarised in Table 2. Collections 
are held in the West Australian Herbarium, Perth. Island 
biogeographical interpretations are given by Buckley (1982). The 
overall flora is a subset of the Barrow Island flora, with the 
addition of six species not recorded from Barrow Island itself: 
Capparis sp. (RB 7155), Crotalaria medicaginea . Dicladanthera 
forrestii . Launaea sarmentosa . Lawrencia sp. (RB 7137), and the 
unidentified sterile species RB 7144 (Dolichandrone sp.?) 



Additional differences from corresponding Barrow Island vegetation 
types include the following. Firstly, the showy pink-flowered 
Ptilotus exaltatus is conspicuous on the higher limestone terraces of 
the Lowendal Islands but extremely scarce on Barrow Island, even in 
directly comparable habitats. Its abundance on the Lowendal Islands, 
which was also noted for the largest island by Serventy and Marshall 
(1964), is perhaps due to the absence of herbivorous macropods, which 
might also account for the greater frequency of Sesuvium 
portulacastrum , Tribulus cistoides , Atriplex semilunaris , Setaria 
dielsii, Commelina ensifolia , Ipomoea pes-caprae , Indigofera trita and 
Portulaca intraterranea on the Lowendal Islands (Buckley 1983). 



Secondly, individuals of Nicotiana occidentalis . Chamaesvce australis . 
Calandrinia balonensis and Portulaca intraterranea on the Lowendal 
Islands are more succulent than those on Barrow Island or the 
mainland: this might be due to greater salinity, lower grazing 
pressure or the greater frequency of seabird colonies with associated 
disturbance and input of nitrogen and phosphorus. 

Thirdly, defoliation of Capparis spinosa by Caper White butterfly 
larvae is more severe on the Lowendal Islands than on Barrow Island or 
the mainland. On such small islands, the relative magnitude of 
fluctuations in the host and herbivore populations may be greater, 
with 1980 perhaps a year when the herbivore/host ratio was high. 

Fourthly, the mangrove patches on island A and one of the patches on 
island C (Fig.1) are now isolated from the sea by extensive sand 
barriers, and the individual plants are all large trees, their trunks 
partially sand-swamped. Since Avicennia marina seedlings establish 
only in tidal mud, this indicates that the sand barriers formed after 
the Avicennia populations established. It is possible that mature 
Avicennia trees cut off from the sea by a sand barrier could still 
propagate from seedlings whilst a mud patch remained around the roots, 
fed with seawater by percolation through the sand, but once the mud 
was swamped with sand further seedling establishment would be very 
unlikely. Hence an age on the mangroves would give a minimum age for 
the sand barriers. Such an age would be obtainable by radiocarbon 
dating if possible anomalies due to the past detonation of atomic 
bombs on the nearby Montebello Islands could be assessed and allowed 
for. 



78 

ACKNOWLEDGEMENTS 

This study was carried out under a Rothmans Research Fellowship at the 
Department of Biogeography and Geomorphology, Australian National 
University, and a Barrow Island Research Grant from West Australian 
Petroleum Pty Ltd and the West Australian Wildlife Authority. Harry 
Butler gave assistance and consultation in the field. Assistance in 
plant identification was given by Harry Butler, the West Australian 
Herbarium and the Herbarium Australiense, and by Prof F.R. Fosberg of 
the Smithsonian Institution. Aerial photographs were provided by the 
West Australian Wildlife Authority. 



REFERENCES 

Buckley, R.C., 1982. The habitat unit model of island biogeography. J. 
Biogeog . , 9, 339-344. 

Buckley, R.C., 1983. The flora and vegetation of Barrow Island. L, 
Roy. Soc. West. Aust. 66, 91-105. 

Hill, F.L. 1955. Notes on the natural history of the Monte Bello 
Islands. Proc. Linn. Soc. Lond . . 165 . 113-124. 

Serventy, D.L., and Marshall, A.J., 1964. A natural history 
reconnaissance of Barrow and Montebello Islands, 1958. CSIRO 
Division of Wildlife Research Tech. Pap. No. £. 



«=>p 



N 



M 






Lowendal 
Islands 



<=© 



G>F 


QK 


Cfc,E 


0J 


£=> 


dG 


*MT 


%H 


k ^\ 





km 2 approx 





<^ 



Fig. 1. The LowendaL Islands and their location 



80 



TABLE 1: BIRDS RECORDED FROM LOWENDAL ISLANDS BY SERVENTY AND 
MARSHALL IN 1958 



Wedge-tailed shearwater 

Pied cormorant 

Reef heron 

Pied goose 

Osprey 

Kestrel 

Pied oyster-catcher 
Sooty oyster-catcher 
Red-capped dotterel 
Caspian tern 
Silver Gull 
Bar-shouldered dove 
Welcome swallow 
Australian pipit 
Yellow silvereye 
Singing honeyeater 
White-breasted wood-swallow 



Puffinus pacificus (Gmelin) 
Phalacrocorax varius (Gmelin) 
Egretta sacra (Gmelin) 
Anseranas semipalmata (Latham) 
Pandion haliaetus (L.) 
Falco cenchroides (Vigors and 

Horsfield) 
Haematopus ostralegus (L.) 
Haematopus fuliginosus (Gould) 
Charadrius alexandrinus (L.) 
Hydroprogne caspia (Pallas) 
Larus novae-hollandiae (Stephens) 
Geopelia humeralis (Temminck) 
Hirundo neoxena (Gould) 
Anthus novae- seelandiae (Gmelin) 
Zosterops lutea (Gould) 
Meliphaga virescens (Vieillot) 
Artamus leucorhynchus (L.) 



31 



TABLE 2: 


PLANT SPECIES RECORDED FROM LOWENDAL ISLANDS IN SEPTEMBER 1980 


SPECIES 


ISLAND: ABCDEFGHJKLM 



Abutilon exonemum F. Muell. * 
Abutilon leucopetalum F. Muell . 
Acanthocarpus preissii Lehm. 
Amaranthus viridis L. * 

Atriplex isatidea Moq. * 

Atriplex semilunaris Aellen * 
Avicennia marina (Forst . )Vierh. * 
Boerhavia mutabilis R.Br. * 

Boerhavia re panda Willd. * 

Calandrinia balonensis Lindl. 
Canavalia rosea (Sw.) DC. 
Capparis spinosa L. * 

Capparis sp. , RB 7155 

Cassytha filiformis L. * 

Chamaesyce sp. 1 * 

Chamaesyce sp. 2 * 

Chamaesyce australis (Boiss . ) Hassall 
Chamaesyce myrtoides (Boiss .) Hassall * 
Chrysopogon fallax S.T. Blake 
Commelina ensifolia Benth. 
Corchorus parviflorus (Benth . ) Domin . 
Dactyloctenium radulans (R. Br) Beauv . 
Dicladanthera forrestii F. Muell. 
Digitaria sp. , RB 7040 
Dysphania plantaginella R.Br. 
Enchylaena tomentosa R.Br. 
Eragrostis basedowii Jed. 
Eulalia fulva (R.Br.) Kuntze 
Euphorbia tannensis Spreng. * 
Ficus platypoda (Miq.)A.Cunn.ex Miq. 
Flaveria australasica Hook. * 
Gomphrena conferta Benth . 
Goodenia microptera F. Muell. 
Indigofera trita L. 
Ipomoea pescaprrae (L.)R.Br. 
Launaea sarmentosa* (Willd.) Alston * 
Lawrencia sp., RB 7137 * 

Lepidium leptopetalum F. Muell. 
Melhania incana Heyne * 

Myoporum acuminatum R.Br. 
Neobassia astrocarpa (F . Muel 1) AJScott 
Nicotiana occidentalis Wheeler * 
Panicum australiense Domin. * 
Pittosporum phillyraeoides D.C. 
Plumbago zeylanica L. * 

Portulaca intraterranea J. M. Black * 



82 



TABLE 2: 


PLANT SPECIES RECORDED FROM LOWENDAL ISLANDS IN SEPTEMBER 1980 


SPECIES 


ISLAND: ABCDEFGHJKLM 



Portulaca pilosa L. 
Ptilotus exaltatus Nees . 
Ptilotus villosiflorus F.Muell. 
Rhagodia obovata Moq . 
Rhynchosia minima (L.) D.C. 
Ruellia primulacea Nees 
Salsola kali L. 
Sarcostemma australe R.Br. 
Scaevola crassifolia Labill. 
Scaevola cunninghamii Labill. 
Scaevola spinescens R.Br. 
Sclerolaena spinosa (Ewart et 

Davies) A.J. Scott 
Sesuvium portulacastrum L. 
Setaria dielsii Herm. 
Solanum esuriale Lindl. 
Sorghum plumosum (R.Br.) Beauv. 
Spinifex longifolius (R.Br.) 
Sporobolus australasicus Domin. 
Sporobolus virginicus (L.) Kunth 
Threlkeldia diffusa R.Br. 
Triodia angusta N.T. Burb. 
Triraphis mollis R.Br. 
iDolichandrone sp. , RB 7144 



NOTES ON A BRIEF VISIT TO SERINGATAPAM ATOLL 
NORTH WEST SHELF, AUSTRALIA 

by B.R. Wilson 



INTRODUCTION 



In October 1978, the author joined the Soviet Research 
Vessel 'Professor Bogorov' during a cruise in 
Northern Australian waters. Three working days were 
spent at Ser ingapatam Atoll on the North West Shelf. 
This reef is of great interest to students of 
Australian coral reefs and marine biogeography because 
of its geographic position and because it is one of 
the very few atolls in Australian waters. There are no 
published records of its biology and little on its 
geology. Therefore, although there was too little 
time for detailed studies, it seems appropriate to 
place on record observations made on the structure of 
the reef and its fauna. 



The expedition was one of a series conducted by the 
Pacific Institute of Bio-organic Chemistry, Far-East 
Science Centre, Vladivostok. The basic objective 
of the expedition was to conduct comparative chemical 
and biochemical studies of marine invertebrates and 
algae inhabiting coral reefs with particular reference 
to some low molecular compounds and biopolymers which 
possess anti-tumoral and anti-fungal activities. 



Museum of Victoria, Melbourne, Australia 

Present address: Dept. of Conservation and Land Management, 
P.O. Box 104, Como 6152, Western Australia 



Atoll Res. Bull. No. 292: 33-100, 1935, 



E 



"3 00 ZZl 
I i o 




85 



REEF STRUCTURE AND FORM 



Position. On the Australian hydrographic chart 
(1047) Seringapatam Atoll is shown with its centre at 
13° 40'S, 122° 5'E. However, the Master of the 
'Professor Bogorov', by use of 'Sputnik' fixes, 
advised that the approx. centre of the lagoon has the 
co-ordinates 121° 59' 35"E, 13° 40' 35"S. 

Shape and Dimensions. Seringapatam Reef is an atoll 
possessing the habitats typical of Indo-West Pacific 
inundated atolls. The atoll is roughly trapezoidal 
with the longest (NE) side being almost straight and 
bearing about 140 . Its length is approximately 4.5 
nautical miles (NW-SE) and its maximum width 
approximately 3 nautical miles (Fig. 1). The 
peripheral reef, which varies from 300-500m wide, 
encloses a broad and deep lagoon. 

Relief and Geology. At high tide no emergent structures 
are visible but at mid and low tide the central part 
of the annular limestone reef emerges. The highest 
zone or reef crest is located about 100m behind the 
outer reef edge and behind that the back-reef varies 
from about 200 to 300m wide. 

Reef flat. Around the entire perimeter of the atoll on 
the reef crest there is a well developed 'boulder zone' 
which provides conspicuous irregular relief to the reef 
when seen from a distance at mid or low tide. Most of 
the boulders are free-lying blocks of reef limestones; 
relatively few are coral slabs. In addition to the 
'loose' boulders there are many 'attached' erosional 
relics, or stacks, which are part of the limestone 
substrate, the highest being about 2m tall (Plate 1). 
Many of them are mushroom-shaped, have a dense calcrete 
capping and are deeply undercut by biological and 
physical erosive forces (Plate 2). Burrowing barnacles 
[LZthotnya. vale.ntla.na) are the principal biological 
erosive agent. Those which have been completely 
undercut have toppled to become reef crest boulders. 



86 



The reef limestone itself is f os s il f erous , containing 
abundant fossil corals and some molluscs. It is assumed 
to be of Pleistocene age. 

Large sand cays exposed at low tide were observed in the 
back-reef zone at the NW and NE corners. There may be 
others at the southern end although none were visible 
when the ship passed around that end of the atoll at low 
to mid tide. 

Channels. A shallow, bent channel was located on the 
NE side at about 122° 0' 15"E, 13° 40"S (Fig. 1). It can 
be entered only by small vessels drawing less than about 
2.5m and only at high to mid tide. At low tide there is 
a torrent of water pouring out of the lagoon making 
extremely turbulent and hazardous conditions. The channel 
is about 80m wide at its seaward entrance and narrows and 
divides as it enters the lagoon where there is a dangerous 
central patch reef which is very difficult to see in the 
late afternoon due to the angle of the sun. The deepest 
and safest arm of the inner channel turns NW . 

There may be other similar channels into the lagoon, 
perhaps at the southern side, but none could be located. 



Reef Front (= fore-reef). This was examined at a number 
of locations along the NE side. There the intertidal 
part of the reef-flat (i.e. seaward of the reef crest to 
the reef-edge) has a width of about 100m. There is no 
raised rim at the reef-edge and no prominent spures and 
grooves or drainage gutters; the reef-flat slopes gently, 
or with a series of little terraces up to 10cm high, to the 
reef-edge (Plate 3). Periodic broad, shallow, depressed 
zones carry the out-flowing water off the ree-flat at low 
tide. The reef-edge itself is very irregular and 
indistinct. There is little living coral or crustose 
coralline algae on the reef-front which has a close- 
cropped cover of leafy brown algae. 

From the intertidal reef-edge there is a broad 
reef-front slope progressing seaward for a distance 
of 100-150m apparently around the entire perimeter 
of the atoll. On the side of the atoll which was 
examined (NE), the upper part of this zone is gently 
sloped with an average angle of about 5 . It is 
characterized by high sub-tidal spurs or ridges roughly 
normal to the reef-edge and bearing moderate growths 
of living scler act inian corals, alternating with deep 






surface 






-10 






_ 20 






_ 30 






_ 40 
_ 50 


2 


^ 


_«0 




I 


_ 70 




I 


_80 






_ 90 

I . , 







surface 




DISTANCE IN METRES 



Fi9. 2 

Small boat sounding transects 1 to 6 
across the reef-front slope normal to 
the edge along the NE side of 
Seringapatam Atoll. 



88 



PLATE CAPTIONS 



Plate 1: Residual reefal limestone stack on the 
reef-crest, NE side of Seringapatam 
Atoll. Note also the numerous 
free-lying boulders. 



Plate 2: Residual reefal limestone stack with 
lithified crust on the reef-crest, NE 
side of Seringapatam Atoll. Note 
extensive burrowing by barnacles on 
the sides. 



Plate 3: The weakly terraced reef-edge at the 
NE side of Seringapatam Atoll at 
low-t ide . 



Plate 4: Valeriy Kiselev sampling 

alcyonarious on the vertical wall, NE 
side of Seringapatam Atoll; 55m. 




+■> 
Ql 



* 




0) 

a. 




CO 

a> 

+■> 

a. 



90 



grooves or gutters which have coa 
boulders and rubble on their beds 
spurs are deeply undercut and cav 
and groove system terminates at a 
12-15m. Below this the slope inc 
Here coral rubble and soft corals 
there are some living scleractini 
slope terminates abruptly at the 
'drop-off' at 30-40m. Deep dives 
wall were made at 4 localities in 
channel (between stations 5 and 3 
80m. Narrow ledges no more than 
only horizontal features of the w 
75m. At that depth a steep rubbl 
observed, but this too was only 1 
that was another vertical wall wh 
sight. There is little scleracti 
the vertical rock face, but numer 
tree-like gorgonians (Plate 4). 



rse sand and coral 

The sides of the 
ernous . The spur 

depth of about 
reases to about 15 . 

predominate but 
ans . The reef-front 
edge of a vertical 

down the drop-off 

the vicinity of the 
) to depths of 60 to 
lm or so wide are the 
all as far down as 
e slope (c_a. 25 ) was 
0-12m wide and below 
ich continued out of 
nian coral growth on 
ous soft corals and 



A series of small-boat sounding transects normal to 
the reef-edge along the NE side confirmed the existence 
of the vertical drop along the whole length of that 
side (Fig. 2). The traces ended abruptly at the 
drop<-off edge because the depth at the foot of the 
vertical wall exceeded the maximum range of the sounder 
used ( 1 10m) . 

No equivalent deep dives or small-boat sounding transects 
were made on the NW, SW and S sides of the atoll and 
it was not determined whether there is a vertical rock 
wall there also. However, 'Bogorov' echo-sounding 
traverses as close as possible, both around and normal 
to the reef (Figs. 1 & 3) show that there is at least 
an extremely steep if not vertical 'slope' around 
the entire perimeter of the atoll down to about 200m. 
Below that there is a steep slope of about 35 to the 
sea floor which is reached at about 500m at the SE end 
and 750m at the NW end (Fig. 3). 



Back-Reef. Behind the reef crest there is a wide 
back-reef zone, rich in living corals in the outer part 
and sloping very gently back to a sandy inner part. At 
its sandy inner edge, at a depth of about 5m, the 
back-reef slope begins, angling abruptly (ca. 20-30 ) 
down to the bottom of the lagoon. On this steep slope 



91 



and on the sandy inner back-reef there are extensive 
dense patches of staghorn Aa/topoh.a corals. 

Lagoon. A small boat sounding traverse was made down 
the centre of the northern part of the lagoon. The 
maximum depth measured was 27m. The bottom is very- 
irregular with numerous coral patch reefs, mounds 
and pinnacles but these rarely rise closer than 2m 
to the surface. (Thus, once entered the lagoon is 
freely navigable by small boats.) 

At a diving station near the centre of the lagoon 
the bottom sediment was found to consist of thick, 
white, calcareous silt. Bottom visibility when 
undisturbed was only 3-^m. Corals were diverse in 
patches; staghorn Kdh.opoK.0. species were the most 
abundant but many foliose corals such as ?achyts2.fii2.i> , 
Ech-Lno phy It-la. , Mzsiullna and massives like Lobophyllia 
were also common. 

DISCUSSION 

Teichart & Fairbridge (19I+8), Jones (1973), Hinz et al 
(1978) and others have discussed the geological history 
of the North West Shelf and Sahul Shelf region. The 
outer part of the shelf is believed to have been 
subjected to substantial subsidence since the Mesozoic. 
Seringapatam, its neighbour Scott Reef and the Rowley 
Shoals further south, rise from the sea floor of the 
depressed continental slope at depths of 500 to 800m. 
Seringapatam and Scott lay on the crests of anticlinal 
trends. Below Scott Reef there is a thickness of 
more than 2000m of Tertiary and Quaternary reefal 
limestone (Jones, 1973). Faulting is common along the 
shelf margin in this region; the fault direction is 
usually NW, ie normal to the margin (Jones, 1973). 



Thus, and taking account of its extremely steep and 
rather straight-sided, trapezoidal form, it seems 
reasonable to interpret Seringapatam as a coral reef 
structure of considerable antiquity, built originally 
(early Tertiary?) upon an upthrust block in the faulted 
basement. The strikingly straight NE side, for example, 
lies in the regional fault direction and implies that 
the shape and character of this atoll are structurally 
controlled. Continuing regional subsidence and rapid 
reef growth since the early Tertiary has resulted in 
the present flat-topped, tower-like structure. The 



92 



vertical sides in the upper 200m or so are probably 
a result of coral growth and terrestrial erosion 
in successive Pleistocene eustatic stages. 

The undercut stacks of the reef-flat boulder zone are 
interpreted as erosional relics of an earlier 
reef-flat which stood about 2m higher than the present 
one. The rate of erosion is very rapid and the age 
of the higher reef-flat was probably Holocene. 



MARINE FAUNA 



Several hours were spent during the afternoons of 
October lU and 15 on the reef-flat at low tide at a 
location on the NE side of the atoll about 0.5km north 
of the channel. Some hand collecting was done there. 
Collections were also made by snorkel and scuba diving 
in the lagoon and along the reef-front slope on the NE 
s ide . 

Voucher specimens of the samples from which the ship's 
biochemists took extracts for analysis were lodged 
at the W.A. Museum for future reference. Other 
specimens of echinoderms, molluscs and some 
scler ac t ini an corals, representative of the common 
elements of the fauna, were also collected and lodged 
at the W.A. Museum. 

Although these collections were far from exhaustive 
they seem sufficient to characterize the invertebrate 
fauna. Figure h shows diagrammatically the reef-flat 
habitats sampled. 

Habitat Zones : 

1. The platform surface, crevices and 
shallow pools of the outer reef-flat. 

2. Under coral and reef-rock slabs of the 
reef crest. 

3. On high rocks and stacks of the high 
intertidal zone on the reef crest. 



93 



k. Sand and rubble substrates of the 
back-reef shallows , 

5. Under coral and reef-rock slabs of the 
back-reef shallows. 

Tables 1 and 2 list the common macro-molluscs and 
echinoderms taken. 

CORALS 

Although no extensive collection of corals was made 
during this expedition the variety observed was great 
in the lagoon where many ramose, foliose, massive and 
encrusting forms occurred. On the other hand, growth 
of hard corals on the reef-front slope seemed less 
luxuriant than on many near-shore reefs further south 
on the Western Australian coast. 

Soft corals were a conspicuous feature of the benthic 
fauna, especially on the lower parts of the reef-front 
slope on the WE side of the atoll where they out- 
numbered scleractinians . In that situation the 
alcionacean genera Mzpthya, AZc-ion-ia , and KLdi.ondh.i.a. 
were especially abundant, as were the gorgonaceans 
VZzxonxfia. and Gotgon-ia. These animals were sampled 
extensively by the biochemists. 

ECHINODERMS 

A representative series of the common echinoids and 
asteroids was retained (Table 2). The material 
identified indicates that the echinoderm fauna is 
typical of the Central Indo-West Pacific region. Many 
holothurians were collected for biochemical analysis 
but not all the specimens were kept for subsequent 
identification. No crinoids and only a few ophiuroids 
were collected. 

MOLLUSCS 

The macro-molluscs of the reef-flat form a community 
typical of oceanic atolls and reefs of the Central 
Indo-West Pacific region with browsing and predatory 
prosobranchs being most conspicuous. The fauna is 



94 



different in several respects to the analogous reefs 
on the coastal islands of the Western Australian coast. 

The common WodtoZui, is M. autiicuZatuA , a very- 
widespread intertidal mussel found on clear-water, 
oceanic reefs. On the turbid-water coastal reefs north 
of North West Cape, this species is absent and instead 
one finds there M. nippontcui Oyama, 1950. 

On the mainland and coastal reefs, coral rocks and 
boulders of the reef-flat are heavily bored by 
Ltthophaga turn* , L. obn^a, L. natuta and L. maZaccana. 
At Seringapatam species of LJ.thopha.ga. were not seen 
and instead the coral boulders of the reef-crest 
(Zone 3) were heavily bored by the cirripede 
Ltthotiya vate.nti.ana. There were no oysters on the high 
intertidal rocks and stacks; the only molluscs collected 
there were VatzZZa &Ze.xuo6a, Ucfiita sp. , and 
Thai* aK.mtgQ.fia. 

Cztiithtum noduZo&um , Zypfiaza ht&tfiio, C. dzpxz66a and 
Lambti chtagfia are all conspicuous gastropods on the 
Seringapatam reef-flat (Zones 1 and 2) but these are 
absent or rare on the northern coastal reefs where the 
water is more turbid. ( Lambtt, chtagfta and 
ModtoZiU) aun.tcuZatu.6 are found on the fringing reef 
south of North West Cape where the water is clear.) 

Faunal differences of this kind are interpreted as 
ecological, the fauna of Seringapatam showing the 
characteristics of an isolated oceanic atoll not 
subjected to turbid coastal water. 

Another noteworthy feature of the fauna is the presence 
of the Pacific thaid Na&6a i>ZHta. On the mainland and 
coastal reefs, including those south of North West Cape, 
the Indian Ocean species Ua&i>a fatianco Z-ina Bruguiere 
occurs but not N. -iztita.. Also, on the coastal reefs, 
the Indian Ocean species Vfiup-tna. Zobata. Blainville 
occurs together with the Pacific V . gh.oiitiU.Za.Kta, 
while at Seringapatam only the latter is found. These 
observations suggest that the intertidal fauna of 
Seringapatam lacks the peculiarly Indian Ocean elements 
in favour of their Pacific analogues. 



95 



DISCUSSION & SUMMARY 

The intertidal invertebrate fauna of Ser ingapatam is 
typical of Indo-West Pacific oceanic atolls and seems 
to have a Pacific rather than Indian Ocean flavour. 
This is not inconsistent with the location and the 
north-easterly ocean currents. 

There is a very diverse scleractinean coral fauna 
although soft-corals dominate the reef-front slope, at 
least along the NE side. 

The reef-flat molluscan fauna is dominated by browsing 
and predatory prosobranchs . Suspensory-feeders are 
unc ommon . 



mean sea level 



lagoon 




transect IN 



transect M 



Fjg.3 : Diagrammatic section through 

Seringapatam Atoll along the NW-SE 
axis, derived from sounding transects 
II and III by the R.V. 'Professor 
Bogorov' and observations on the reef. 



o> 



Hh 




o 


<u 


<«H 


43 


Cm O 


P 


V 


« 


0) >>- 


-> C 


M -P 


H -H 


•rH 




CD C 


• T3 


a: -h 


bO (1) 


-p o 


•H ,£> 


•rl 


Pn -H 


W > 


M 


w 


0) O 


o cu 


o> co 


m .c 


CO 01 


O P 


^ T3 


cd 




B 


LT\ CO 


C -H 


<1J 


o 


P C 


•H H 


o o 


-P H 


<D tsl 


o o 


CO 


<D -P 


C -P 


[0 < 


cd cd 




M P 


O £ 


P -H 


•H cd 


43 


-p p 


P cd 


cd cd 


cd X! 


6 pu 


O 


6 cd 


,0 bC 


cd W) 


G 


M C 


H -H • 


t>D -H 


r-\ > p 


cd « 


cd O X 


•H <U 


S fl D 



Q c/3 co to -p 



I I 



4 4 



97 



TABLE 1 Common molluscs of the reef flat in the vicinity 
of small-boat transect 5. See Figure 4 and text 
for explanation of zones. 

ZONES 



MOLLUSCS 



12 3 4 5 



HodloluA autiJ-CU.Za.tuA Krauss, 1848 x 

Th.ldac.Yia maxima RBding, 1798 x 

Tn.ldacna AquamoAa Lamarck, 1819 x 

HlppopuA hlppopuA (Linnaeus, 1758) x 

Vitaqum ln.aau.rn (Linnaeus, 1758) x 

\kanman oa toma chnyAtomuA (Linnaeus, 1758) x x 

TfiochuA maculatuA Linnaeus, 1758 x 

TfiochuA pqh.ami.Ji Born, 1778 x x x 

Patzlla {IZXUOAa Quoy & Gaimard, 1834 x 

Ne.fii.ta sp. x 

Czn.ltb.lum noduloAum Bruguiere, 1792 x x 

Czfii.tk-iu.rn zchlnatum Lamarck, 1822 x 

Rhino clav Ia sp. x 

Sth.ombuA IzntlglnoAa Linnaeus, 1758 x 

LamblA chlagfia Linnaeus, 1758 x 

Cypfiaza capu.tAZfip2.ntlA Linnaeus, 1758 x x 

Cypfiaza hlAtfilo Gmelin, 179 1 x 

Cypftaza dzpfiZAAa Gray, 1824 x 

Cypfiaza lynx Linnaeus, 1758 x 

Cypfiaza lAabzlla Linnaeus, 1758 x 

Cypfiaza tlgfilA Linnaeus, 1758 x 

Cypfiaza monzta Linnaeus, 1758 x 

BufiAa bufionla Gmelin, 179 1 x 

BafiAa Cfiuzntata Sowerby, 1835 x 

HaAAa AZfita Bruguiere, 1789 x 

VZfilAtZfinla naAAatula Lamarck, 1822 x 

\JaAum czn.amlcu.rn Linnaeus, 1758 x 

\JaAum tuhblnzllum Linnaeus, 1758 x 

Latlnolagzna Amahagdula Linnaeus, 1758 x 

ThalA anmlgzna Link, 1807 x 

TnalA sp. x 

Vfiuplna gnoAAulanla RBding, 1798 x x 

Vhuplna n.lclnUA Linnaeus, 1758 x 

ChlconzuA bnunnzuA (Link, 1807) x 

ConuA llvlduA Hwass, 1792 x 

ConuA filavlduA Lamarck, 18 10 x 

ConuA mllZA Linnaeus, 1758 x x 

ConuA hattuA Hwass, 1792 x 

ConuA ImpZKlallA Linnaeus, 1758 x 

ConuA manmohZuA Linnaeus, 1758 x 

ConuA coh.onatuA Gmelin, 179 1 x 

ConuA AponAallA Hwass, 1792 x 



98 

TABLE 2 Common echinoderms of the reef flat in the 

vicinity of small-boat transect 5. See Figure 
4 and text for explanation of zones. 

ZONES 



ECHINODERMS 



12 3 4 5 



Pafia.6aZe.nia gfiatio6a A. Agassiz, 1863; 

burrowing in the outer surface of 

Tfiidacna maxima x 

Echinomztfia mathazi (de Blainville, 1825); 

burrowing in hard reef surface x 

Echinothfiix diadzma (Linnaeus, 1758); 

pools x 

TftipnZU6tZ6 gfiatilla (Linnaeus, 1758); 

pools x 

Eucidafii6 mztutafiia (Lamarck, 1816) x 

Echinonzu6 cyclo6tomu6 Leske, 1778; 

buried in sand under stones x 

Cu.lc.ita novae.gu.inaz Muller & Troschel, 1842; 

pools x 

Linkia laevigata (Linnaeus, 1758); 

pools and reef surface x 

Linkia mutti&ofia (Lamarck, 18 16) x 

0phida6tZK gfiani^Zfia Lutken, 1872 x 

A6tZfiina ZZphzu6 (Muller & Troschell, 1842) x 

A6te.fie.pAiA cafiini^e-fia (Lamarck, 1816) x 

Uafidoa tu.be.ficu.lata Gray, 1840 x 

Le.hina6te.fi luzonica (Gray, 1840) x 

Stichopu.6 cklofLOnotu6 Brandt, 1835 x 

Thzlonota ananu6 (Jaegar, 1833) 
Bohad6cnia an.gu6 Jaegar, 1833 

OphiaH.th.fLum pictum Muller & Troschell, 1842 
Ophiocoma dzode.file.ini de Lorid, 1899 
Ophian.ac.hna incfta66ata (Lamarck, 18 16) x 



99 

ACKNOWLEDGEMENTS 

While aboard the 'Professor Bogorov' I was treated with 
extraordinary hospitality and given much help in 
pursuing my interests. I sincerely thank Captain Gennady 
Nozdrin, and the Expedition Chief, Dr Valeriy Rasskazov 
for their hospitality and good fellowship and all those 
members of the crew who made the expedition such a 
memorable one for me. In particular, I wish to 
acknowledge the friendship and assistance of my diving 
'buddy', the late Valeriy Kiselev who drowned recently 
in the North Pacific. 

I am especially grateful to Mrs L.M. Marsh of the 
Western Australian Museum and Dr F. Rowe of the 
Australian Museum for identifying the echinodernus . 



REFERENCES 

Hinz, K., Beiersdorf, H., Exon, N.F., Roeser, H.M., 

Staff, H.M.J. , and Stackelberg, U. von (1978) 
Geoscient if ic investigations from the Scott 
Plateau off north-west Australia to the Java 
Trench. B.M.R. J. Aust . Geol . & Geophys . 
3_: 319-340. 

Jones, H.A. (1978). Marine Geology of the North-West 
Australian Continental Shelf. B.M.R. Bull . 
136 : 1-102, 2 maps . 

Teichert, C. and Fairbridge, R.W. (1948). Some Coral 
Reefs of the Sahul Shelf. Geogr. Rev . 38 : 
222-249. — 



SEA SNAKES COLLECTED AT CHESTERFIELD REEFS, CORAL SEA 
by Sherman A. Minton and William W. Dunsonl 



INTRODUCTION 

This report describes a collection of sea snakes made at the 
Chesterfield Reefs June 19-25, 1981. We know of no previous collection 
of sea snakes from this locality. The Chesterfields are a group of reefs 
and sand cays located in the southern part of the Coral Sea about 945 km 
ENE of Rockhampton, Queensland and about 630 km almost due west of the 
northern tip of New Caledonia (Fig. 1). Troughs with depths of 1000-3000 
m separate them from New Caledonia and the Great Barrier Reef complex. 
Loop Islet at the southern tip of the group (Fig. 2) lies at 19° 57' south 
and 158° 28' east. 

METHODS 

Most snakes were captured in nets while snorkling or scuba diving at 
depths of 20 m or less. Underwater visibility was generally good, and 
water temperatures 22-25°C. A few snakes were netted from small boats or 
from the R/V Acheron or were found stranded on sand cays. A total of 79 
snakes was collected and 34 preserved. These have been deposited in the 
Field Museum of Natural History, Chicago, and the Australian Museum, 
Sydney. Six species were identified in the material collected. 

RESULTS 

Acalptophis peronii (4 collected; 2 preserved) . All specimens were 
taken over relatively flat, open areas of sand at depths of about 10-15 m. 
The smallest had a total length of 45 cm, the largest measured about 85 cm. 

Aipysurus duboisii (2 collected and preserved) . One specimen was 
collected in a narrow coral passage at a depth of 14 m; the other was found 
stranded and dead on Anchorage Islet. Total lengths were 83 and 74 cm. 



Department of Microbiology and Immunology, Indiana University School of 
Medicine, 1100 West Michigan St., Indianapolis, IN 46223, U.S.A., and 
Department of Biology, The Pennsylvania State University, 208 Mueller 
Laboratory, University Park, PA 16802, U.S.A. 

Atoll Res. Bull. No. 292: 101-108, 1985. 



102 



Aipysurus laevis (2 collected; 1 preserved). One taken on the side 
of a bommie at a depth of about 6 m; the other over open sand at about 
15 m. Both were small adults about 1 in in total length. 

Emydocephalus annulatus (15 collected; 7 preserved). Five were 
found over open sand, the others in coral passages or around bommies. 
Depths varied from 3 to 15 m. One large snake was probing crevices in a 
bommie while two small fish seemed to be harassing it by biting at its 
neck. It reacted by shaking them off occasionally. After about 10 
minutes the snake glided into deeper water and continued to explore coral 
rubble; the fish did not follow. Another snake had its head and forebody 
in a hole in the sand. A courting pair of snakes were observed June 21. 
Three juveniles 31-34 cm in total length were taken. The smallest of 
these was lying dead on the bottom. Five of 6 preserved adults were 
uniformly black; one large female showed numerous cream-colored scales 
irregularly scattered. Two adults of the ringed pattern morph were seen. 

Hydrophis sp. (40 collected; 20 preserved). This small-headed sea 
snake was plentiful off Loop Islet in an area of flat, open sand lightly 
covered with filamentous algae at a depth of 7-11 m. The snakes were 
usually seen lying on the bottom often with the head and forebody buried 
in sand or were swimming slowly a few centimeters above the bottom. Five 
were collected by dip-netting from the ship while anchored off Passage 
Islet at night and others were seen in the illuminated zone but escaped. 
Several of the snakes were juveniles 40-50 cm long. No food was disgorged 
by newly captured snakes, but the digestive tract of one adult contained 
a brown, pasty, homogeneous material. These snakes are similar to 
specimens collected at reefs of the Sahul Shelf and referred to Hydrophis 
melanocephalus (Cogger, 1975). They also fit the description of H. 
melanocephalus from Fiji (Guinea, 1981) and probably are identical with 
the species taken at Saumarez Reef and identified as Microcephalophis 
gracilis (Heatwole, 1975) . However, they appear to differ significantly 
from both melanocephalus and gracilis from Asian waters and probably 
represent an undescribed taxon whose status is currently under 
investigation. 

Pelamis platurus (11 collected; 2 preserved). Eight of these snakes 
were collected June 23 in a slick on the lagoonal side of the inlet north 
of Passage Islet. The others were found stranded on Loop and Anchorage 
Islets. One was seen at night near the ship. Desiccated remains of 5 
other Pelamis were found on beaches. One had been incorporated into the 
nest of a booby. Four of the snakes collected were juveniles about 35- 
40 cm in total length. Two of the others were very dark, large individuals, 
one female having a total length 98 cm. 

We presume all the species are residents of the Chesterfields rather 
than strays, for multiple individuals of all were collected. Collection 
of juveniles of 4 species suggests local breeding. 



103 



DISCUSSION 

Table 1 summarizes information of sea snake distribution along the 
Queensland Coast, the southern part of the Great Barrier Reef, the 
Chesterfield Reefs, New Caledonia and the Loyalty Islands, and Fiji. 
Species diversity is greatest along the Queensland coast probably because 
species with preference for turbid water and a muddy bottom such as 
Aipysurus eydouxi , Lapemis hardwickii , Enhydrina schistosa, and Hydrophis 
elegans find favorable habitat here but not on the Great Barrier Reef or 
eastward. The first three species have extensive ranges indicating good 
powers of dispersal. Mud and turbidity may largely exclude Emydocephalus 
annulatus , a snake of clear water and coral reefs, from the coastal zone. 
Aipysurus laevis and A_^_ duboisii are characteristic coral reef species 
that also are locally plentiful along the Queensland coast. Pelamis 
platurus is not known to breed along the Queensland coast, although beach- 
washed individuals are encountered regularly. It is uncommon on the Great 
Barrier Reef but appears to have a well-established population in the 
Chesterfields. Laticauda colubrina and L^ laticaudata are reported to be 
common in waters around New Caledonia and the Loyalty Islands and also 
occur at Fiji. They are unknown from the Great Barrier Reef and recorded 
from the eastern coast of Australia only very rarely. They appear to 
have reached New Caledonia from the northwest by dispersal along the New 
Hebrides Ridge. Availability of suitable daytime resting and seasonal 
nesting areas may be vital factors influencing distribution of these 
oviparous sea snakes. Astrotia stokesii and Hydrophis ornatus are wide- 
ranging and presumably eurytopic species that may eventually be found in 
the Chesterfields. 

ACKNOWLEDGMENTS 

Supported by NSF grant PCM 79-18393 to W.A.D. We are grateful to 
the University of New England, Armidale, N.S.W., for serving as the 
official sponsor of the expedition. Dr. Harold Heatwole's assistance, 
advice, and enthusiastic participation in the expedition contributed 
greatly to its success. We also appreciate the help of Glen Burns, Glenn 
Stokes, and David Berman in catching snakes. 

LITERATURE CITED 

Dunson, W.A. 1975. Sea snakes of tropical Queensland between 18° and 20° 
south latitude. Tn W.A. Dunson (ed.) The Biology of Sea Snakes , 
Baltimore, University Park Press, pp. 151-162. 

Cogger, H.G. 1975. Sea snakes of Australia and New Guinea. In Dunson, 
op . cit . , pp. 114-115. 

Forne, F. 1888. Sur un cas de mort par morsure du serpent de mer. J. 
Offic. N. Caledonie et Depend. , no. 1520, pp. 335-341 (not seen) 

Gail, R. and Rageau, J. 1958. Introduction a 1' etude des serpents marins 
en Nouvelle-Caledonia. Bull. Soc. Path. Exot. 51: 448-459. 



104 



Guinea, M.L. 1981. The sea snakes of Fiji. Fourth Internat. Coral Reef 
Symp. 

Halstead, B.W. 1970. Poisonous and Venomous Marine Animals of the World, 
Washington, D.C., Government Printing Office, vol. Ill, pi. xxxii. 

Heatwole, H. 1975. Sea snakes found on reefs in the southern Coral Sea 
(Saumarez, Swains, Cato Island). I_n Dunson, op. cit . , pp. 161-171. 

Limpus, C.J. 1975. Coastal sea snakes of subtropical Queensland waters 
(23° to 28° south latitude). In Dunson, op. cit. , pp. 173-182. 

Roux, J. 1913. Les Reptiles de la Nouvelle-Caledonie et des iles Loyalty. 
In Sarasin and Roux, Nova Caledonia , Wiesbaden, pp. 76-160 (not seen) 

Smith, M.A. 1926. A Monograph of the Sea Snakes . London, Taylor and 
Francis. 



K WHITSUNDAY IS. 



S !ci«,-'' SA UMAREZ 
REEFS | REEF 



ROCKHAMpTX) 



CAPRICORN 




CHESTERFIELD 

CROUP/"^ 



GROUP 



CATO IS. 



FRASER IS. 



QUEENSLAND 



BRISBANE} 



N 




NEW CALEDONIA 



Fig. 1. Location of Chesterfield group with respect to New Caledonia 

and the Queensland coast. Dashed line marks approximate outer 
limit of Great Barrier Reef complex. 



Fig. 2. Chesterfield Group showing localities mentioned in text. 
Figures indicate water depths in fathoms. 



LONG ISLAND PASSAGE 




ANCHORAGE 



METERS 1000 



SOUTH ELBOW 



106 



B3FpriB3T:jEX EpnBDTUB^ XJ 



FUT^qn^oo epriBoxriBi & x x 



snanaB^d STUiB-pd 3 3 X X 



TTipxMpaeq sxuisdBT 



¥¥ 2 *ds sjqdoapAH 



¥ j 'ds STqdoapAH ^ X 



snaBuao sxqdoapAH 



sub2bjs sxqdoapAH 



Bsoasxqos FUTjpAqua X 



sn^B^nuuB sn-[BqdaoopXui3 oi XX X 

jo fern paxaasTQ x <^- 



tt§ut^ BaxsasTQ 



TTsa^oqs Bxaoaasv x x 

Erf ' 

PQ 

< STA3BT snansXdTV x x x x 



txnopAa snansXdxv x 



TTSToqnp snansAdxy 



tiuojsd sxqdoadATPDv 





X 






X 


X 






X 






QZ 




CD 

o 


X 


X 




— 


c* 




o ... 




-a 


.i 








CO 






4-1 in 






4-1 




CO 




u 






I — 




en 


cfl 




TS 


• 


4-1 






• o 




c 


01 


S~K 


£ 


en 


cn 






CO —i 




•H 


- 


CO 


Cfl 


i— i 


— 




/-y 


H-l 




cfl 


*w> 


c 


.—1 




cfl 




•—t 


M 




S 




o 


CO 


>> ■x 




00 


>- c 




C/j 


■ 


•H 


l— 1 


4-1 






a> 


a> o 






CO 


4-1 




.—1 


mm 




•— i 


CO CO 




* 


i-i 


Cfl 


T3 


CO 


OO 






CO C 




14-1 




> 


C 


^ 


cn 




r. 


U 3 




CD 


o 


u 


Cfl 


o 


c* 




CO 


&< a 




CD 


4-1 


01 




hJ 


i — i 




01 


>>_^ 




oi 


Cfl 


co 


CO 






^-. 


c 








U 


J2 


<*- 


w 


M 


^c 


•H 


4-1 




(J 




o 


0) 


C 


a 


CNI 


3 


CO CO 




01 


M 




01 


Cfl 


cfl 


o-. 


O 


co i— i 




•H 


CO 


i-l 


a: 




j 


f— < 


•^s 


o 


s- s 


1-4 


M-i 


cfl 




cfl 


01! 






u ^ 


in 


J-l 


0) 


C 


-o 


•H 


- 


1 


CO 


o 


(^ 


Cfl 


•.. 


o 


i—i 


C 


^ 


~~ 


T3 


-a o 


c^ 


S3 


a; 


CO 


0) 


O 




4-1 


c 


c u 


. — 1 






(-1 


•f-l 


XJ 


-a 


•r-l 


cfl 


CO ^3 




C 


N 


0) 


U-J 


o 


= 


£ 


i-H 


— c 


A 


u 


01 


D. 


U 


. — ■ 


cfl 


CO 


CO 


CO -H 


CO 


01 


S-i 




2J 


cfl 






M 


C J= 


D 


X 


CO 


. n 


4J 


o 


t— i 


• M 




0) u 


D. 


4-1 


H 


LPi 


en 




•H 


c; 


•rH 


o) c 


H 





- 


r^ 


0/ 


3 


cfl 


r^ 


• |— 1 


D -H 


-H 


a 


CO 


CT> 


-~ 


a 


O 


CT- 


■H 


O" X 


_: 


C/j 


C/l 


— i 


o 


^ 


x — ' 


— < 


Uh 



107 



















QJ 


















QJ 








QJ 
















X 








43 


#s 














4-> 




)H 




QJ 


3 en 

cfl 3 




. 






• 




>i 




o 




> 


QJ 4-> 




CO 






co 




.-1 




MH 




cfl 


bO CO 




3 






3 




43 








X 


Cfl rH 




QJ 






o 




cfl 




^-N 






P=5 rH 




n 






•rl 




43 




LTl 




QJ 


QJ 




3 






4J 




O 




r^. 




CO 


T3 CJ 










CO 




)-i 




(3^ 




cu 


3 O 




>, 






rH 




ft 




1 — 1 




43 


cfl 




X 






3 








^^ 




H 


CD 




Sh 






ft 














rH 3 




3 






o 




• 




3 






•H 4-1 




QJ 






ft 




en 

3 




o 
en 




en 


cfl cfl 
O 3 




3 






60 




rH 




3 




rl 


^ rl 




XI 






3 




cO 




3 




Q) 


O 




3 






•H 




43 




n 




4-1 


en 






3 






X) 




3. 








cfl 


•H 


CO 










oj 




QJ 




0) 




& 


rH 


•H 




CO 






OJ 




CJ 




0) 






•rl 


X 




X) 






Sh 




O 




CO 




cfl 


O 


ft 




3 




• 


42 




3 








•H 


Cfl 


o 




3 




CO 






cfl 








3 


U 


U 




rH 




qj 


4-J 




rH 




• 




O 


60 


T3 




CO 




n 


o 




QJ 




3 




•3 


>i 




H 




3 


s 




e 




>H 

QJ 




Q) 

rH 


CD 

•H 


K 




X" 




0) 


* 




CO 


• 


4-1 




cfl 


X 


X) 




rH 




xi 


CO 




•rl 


,^v 


4-1 




O 


ft 


3 




QJ 




4J 


rH 




43 


m 


Cfl 






O 


cd 




•H 






CO 




ft 


r^ 


ft 




£ 


rH 






MH 




3 


3 




o 


a> 






QJ 


Cfl 


•> 




V4 




•rl 


XI 
•H 




■3 


r-H 


QJ 
4-1 




23 


ft 


S 




QJ 
4J 


,^^ 


60 


> 




!^ 




cO 




3 


QJ 


cfl 




CO 


X) 


3 


•rl 




ffl 


QJ 


3 




•H 


a 


QJ 




QJ 


QJ 


•rl 


XI 






rH 


U 






o 


60 




X 


3 


x) 


3 







O 


O 




U 


u 


CO 




c_> 


c 


a) 


•H 




4-1 


£ 






3 


a 


Pi 






•H 


a) 








4-1 


.3 




O 


•H 






OJ 


4-1 


i-i 


!>. 




XI 


Cfl 


4-1 




CJ 


S 


XI 




.3 


3 


43 


CO 




QJ 


QJ 


•rl 




o 


3 




4-1 


O 




U 




rl 


ffl 


& 






T3 CO 






o 


>> 


4-> 




H 








en 


3 




3 


^^ 


rH 


CO 




QJ 


!=s 


CO 




QJ 


cO <-\ 




•rl 




4=> 






14-4 


X 


•H 




^ 


•H 






T— 1 


3 


3 




QJ 




X 




cO 


CO 


cfl 




CO 




& 


o 




u 


CO 


ft 




3 


3 


CJ 




QJ 


d) 


3 




• 




•H 


o 




en 


QJ 






M 


^H 


en 


13 


-a 


^ 


^H 


u 






O 


>^ 




3 


43 


QJ 


aj 


aj 


■H 


•H 


X) 




cfl 


CD 


-Q 




3 


ca 


^4 


CO 


3 


4-> 


o 


^ 




QJ 


QJ 






CO 


H 


ft 


CO 


•H 


3 


cfl 


m 




en 


rH 


X) 


• 








43 


e 


QJ 


Sh 








3 


QJ 


^-N 


3 




XI 




u 


H 


60 


XI 




QJ 


t-i 


4J 


3 


QJ 




3 


^ 


qj 


M 




QJ 




M 


QJ 


•H 


cfl 


CO 




cO 


iH 


4J 


3 


CO 


Xl 




•H 


cfl 


O 


QJ 








X 


qj 


a 


•rl 


cfl 




rH 


CJ 




60 


MH 




.H 


cfl 


•3 




X 


QJ 




1 


A 


cO 







3 


6 


3 


3 


ft 


43 




en 


en 


ro 


Pi 






M— i 


3 


3 


O 


O 


1 




•H 


•H 


r-i 




QJ 




•H 


CO 




•H 


rH 


tH 




X 


X 


o> 


XI 


CJ 




■U 


OJ 


CO 


4-> 


cfl 


iH 




ft 


ft 


rH 


3 


3 




3 


u 


QJ 


cfl 


X 


CO 




O 


O 




cfl 


3 




QJ 


ft 


Sh 


iH 


ft 


B 




!h 


U 


A 




X) 




iH 




CO 


3 


QJ 


CO 


• 


T3 


-o 


X 


^-{ 


3 




ft 


CO 




3. 


CJ 




,^N 


f^ 


>, 


3 


•H 


3 






-a 


OJ 


O 





XI 


m 


ts 


pa 


O 


cd 


rQ 




>, 


M 


43 


ft 


u 


QJ 


r- 




^ 


O 


3 




rH 


o 


4-1 




u 


43 


O^ 


4H 


en ^-^ 








tH 


o 




MH 


•H 


•H 


rH 


O 


cd 


>^ 


XI 




CO 


OJ 


3 


rH 


s 


!-4 


V- ' 




!h 


X 


3 




CJ 


pi 


•H 


0) 




O 




en 


T3 O 




3 




O 






.3 


o 


CO 


rl 


QJ 


QJ •!-) 


X) 






rH 




CD 


C/3 


4J 


QJ 


QJ 


•H 


•H cfl 


OJ 


3 






■ 


3 






Xl 


60 


a 


C4H B 


4-1 


O 




4J 


CO 


4-1 


rH 


13 


3 


bO 


QJ 


•H 


•rl 


■rl 




CO 


OJ 


cO 


3 


OJ 


3 


O 


ft 


4-1 CO 


CJ 


4-1 




CO 


u 


4-1 


J! 


i-i 




C_> 


CO 


3 Sh 




3 




a; 


cO 


CO 


cO 


rl 


>. 






QJ -H 


»N 


43 




rH 






CO 


QJ 


rH 


O 


O 


XI QJ 


00 


•H 






ai 


4-1 




<4H 


4-1 


en 


& 


•H 4-1 


00 


r( 




4-1 


43 


3 


o 


QJ 


3 


rH 


4-1 


CO 


00 


4-1 




CO 


4-J 


42 


4-1 


u 


QJ 

)H 


cfl 


4-1 


rH n 


1 — 1 


CO 

•rl 




CO 


3 


4-1 


H 


CO 


cfl 


• n 


CD 


CO 


n 


n 




0) 


•H 


3 


cfl 


QJ 


ft 


QJ 


Cfl 


3 - 


QJ 






•H 




a) 


rH 


•H 


ft 


rl 


QJ 


O /-v 


3 






CJ 


0) 


en 


•H 


o 


cfl 


3 


rH 


•H 00 


!h 


• 




OJ 


u 


QJ 


S 


QJ 




60 




M LO 


O 


rH 




ft 


CO 


!-i 


•H 


ft 


3 


•H 


4-1 


Cfl C3> 


ft 






CO 


Pi 


P-i 


C/3 


CO 


<ti 


MH 


< 


> rH 


^— * 


QJ 
rH 
43 




X 


Pi 


ft 


■X 




-K 




-K 






3 














■K 




c— 








H 



THE UNDERWATER MORPHOLOGY OF PALMERSTON AND SUWARROW ATOLLS 

A 



by J.Irwin 



ABSTRACT 

Methods used and results of echo sounding surveys of Palmerston 
and Suwarrow lagoons, Northern Cook Islands, are given in this paper. 
Notes on the compilation of a bathymetric chart of Palmerston Atoll 
are given and features of the underwater morphology of each lagoon are 
described and illustrated. 

INTRODUCTION 

During September 1981 Palmerston and Suwarrow Atolls in the North- 
ern Cook Islands were surveyed by echo sounder. This work was part of 
a joint N. Z. Oceanographic Institute and Royal Society of London cruise 
using the New Zealand Research Vessel R. V. Tangaroa. Sounding coverage 
was carried out from 5-14 September at Palmerston Atoll but only two 
days, 20 and 21 September were available for soundings at Suwarrow, al- 
lowing only sketch coverage to be made. This note describes the methods 
and results of the sounding survey and data on water characteristics at 
the time of survey. 

Methods and Equipment : 

A 5.5 m aluminium outboard-motor-powered boat and Raytheon survey 
echo sounder (Model DE 719B) with the transducer mounted overside were 
used. Additional and comparative echo soundings were made with a 
Furuno F850 echo sounder. Five stations at Palmerston Atoll and one 
at Suwarrow Atoll were occupied to collect water samples with National 
Institute of Oceanography water bottles for salinity readings and water 
temperatures throughout the water column were recorded using a bathy- 
thermograph. Data collected was used to correct echo soundings for 



* New Zealand Oceanographic Institute, DSIR, P. 0. Box 12-346, 
Wellington North, New Zealand 

Atoll Res. Bull. No. 292: 109-113, illustrations, 1985. 



110 

regional variations to the velocity of sound in water. Foxboro tide 
gauge stations were established at Home and Primrose Islands at Palmer- 
ston Atoll and at Anchorage Island and Suwarrow Atoll, all inside the 
lagoon. Tidal readings from Home Island were used to reduce soundings 
to a common datum, near low water. This gauge and the gauge at Suwarrow 
Atoll were referenced to bench marks. 

Field plots were made using Atoll outlines from 1974 aerial pho- 
tography mapped by photogrammetric methods. Original aerial photographs 
were also available. Sounding traverses were made between known points 
at constant boat speed. The low relief and large width of the lagoon 
made navigation difficult. To aid position fixing and shorten traverse 
lines two large inflatable irridescent red buoys were positioned in an 
approximate N-S line near the centre of the lagoon. These buoys aided 
position fixing, shortened traverse lines and provided end points for 
sounding traverses. Traverse end points were established at atoll 
islands, identified on aerial photographs, or by compass bearings of 
positions inside the atoll edge. Compass bearings were used on long 
traverses to check position along the line. Navigation Chart (B.A. 
1147 Suvorov Islands) was used to plot sounding traverses at Suwarrow 
Atoll. Positions plotted on this chart showed the south west side of 
the atoll further northwards than depicted, by approximately 0.2 to 0.7 
nautical miles, this was confirmed by ships radar from R. V. Tangaroa 
anchored inside the lagoon. Figure 3 shows Suwarrow atoll from 1974 
aerial photography. 

Fifty two traverses approximately 400 metres apart gave coverage 
of Palmerston Atoll for which a bathymetric chart has been prepared for 
publication (Irwin and Main 1983). Ten echo sounding traverses pro- 
vided sketch coverage of Suwarrow Atoll. 

B.A. Chart 1147 Suvorov Islands shows soundings from surveys of 
1900 and 1920; these spot soundings were made before the advent of the 
echo sounder. The present survey provides continuous sounding traverses 
depicting bottom configuration, and in the case of Palmerston Island 
provides new information. 

PALMERSTON ATOLL 

Situated 500 kilometres NW of Rarotonga, Palmerston Atoll takes 
the form of a diamond. Groups of islands are situated about the reef 
concentrated on the N, S, E and W points. Land area is approximately 
400 hectares. The lagoon measures 9 km N-S and 6.5 E-W and the ex- 
posed reef averages 0.5 km across. Several boat passages across the 
reef are situated on the NW side. 

Underwater Morpholog y: 

The underwater morphology is best shown by a bathymetric chart , 
which means depth information from the echo sounder graphs had to be 
plotted on a collector sheet. The extreme undulating bottom allowed 
only high and low points to be read off the graphs along each traverse. 



Ill 

A total of 8870 depths from 52 traverses, half highs and half lows 
were read off, corrected, reduced and plotted on an enlarged outline 
(scale 1:5,000) of the atoll. Final publication scale is 1:18,000. 
The volume of information on any one traverse made it impossible to 
show all the information in chart form. The great variability in 
depth of coral heads (highs) made the drawing of isobaths impossible. 
Consideration was given to showing the heads by symbol along each 
traverse, but their numbers precluded this. The lows, or areas between 
the coral heads, were contoured to show the bottom shape of the lagoon 
(Fig. 2), and this data is shown on the bathymetric chart of the atoll 
(Irwin & Main 1983) . 

Within the lagoon coral heads rise to the surface close to the 
SW shore but few do over the main body of the lagoon. 

Echo sounder records showed the bottom to be covered with coral 
heads. Since they were evident on every sounding traverse, which gave 
good even coverage of the lagoon, the assumption can be made that this 
very high concentration of coral heads covers the lagoon floor. Figure 
1 shows sample echo sounder records at selected positions across the 
lagoon. The Raytheon survey sounder used operated on a frequency of 
200 kHz and has a transducer beam width of 10°. The high sounding rate 
on the scales used for the survey, 534 and 267 soundings per minute, 
with fast graph speed through the machine provided high resolution 
records. Simultaneous soundings using the Raytheon sounder with a 
Furuno model F850 sounder which has a lower sounding rate of 155 sound- 
ings per minute confirmed the Raytheon's superior resolution in these 
conditions . 

The concentration of coral heads appear to be fairly uniform over 
the atoll basin to the deepest (30 + m) areas. Slightly higher concen- 
trations of heads occur in the shallower areas, particularly the N end 
of Palmerston atoll. The height of the heads above the general atoll 
bottom is highly variable. The sounder did resolve small but definite 
flat sandy areas between the coral heads which were confirmed by first- 
hand observations and sampling by divers. 

The inner edge of the reef is steep sided, the 5 m contour falling 
close to the inner reef edge. The 20 m contour also lies close to the 
inner reef edge except in the N sector which is shallower. The area 
within the 26 m contour, about 4 km x 3 km is comparatively flat, slop- 
ing to a low area 1.5 km x 1.2 km within the 30 m contour of similar 
shape to that of the atoll. This contains the deepest recorded depth 
of 34.6 m located to the S and E of the centre of the atoll. Isolated 
highs shown in comparatively deep water appear anomalous but these 
represent large coral head complexes which lie close to the sounding 
traverses. 

Wa ter Ch aracteristics and Tid al Meas ur ements : 

Tempera t ure a nd Salinity 

Measurements were made at Palmerston Atoll on 11 September 1981 
at 5 stations. Water temperature varied less than 0.5°C at any one 



112 

depth, at any sampling position, and less than 0.8°C from surface to 
bottom. Average surface temperature was 26.5 C C decreasing to 26.2°C 
at 10 m, 25.8°C at 16 m, 25.7°C at 22 m and to the bottom at 30 m. 
Surface salinity was 35.47 /°° increasing to 35.50 l°° at 5 m, 35.54 /oo 
at 10 m, 35. 62°/°° at 20 m and 35.66 at 25 m. 

Tidal measurements 

Measurements were made continuously from 4-17 September 1981 at 
Home Island inside the reef. Semi-diurnal tides recorded a maximum 
range of 0.51 m and a minimum range of 0.23 m over the period. The 
gauge on Primrose Island also inside the reef recorded a maximum range 
of 0.41 m and a minimum range of 0.25 m. 

SUWARROW ATOLL 

Suwarrow Atoll lies 950 kilometres NNW of Rarotonga. The atoll 
is near circular in shape with protrusions on the north and east sides, 
an entrance to the lagoon is located on the north-east side. Islands 
are situated around the reef except on the south west side. The 
lagoon is 15 km across E-W and 12 km N-S, the reef averages 0.5 km 
in width (Fig. 3) . 

Underwater Morphology : 

The 10 sounding traverses run provide a sketch survey but the 
complex nature of the atoll with many reefs made drawing a bathymetric 
chart impractical. Figure 3 shows the atoll with sounding traverses, 
and selected sounder records are shown in Figure 4. 

Suwarrow Atoll contains a number of reefs (up to 0.5 km long) 
which are exposed at low water. Coral patches and heads lie in shallow 
water areas close to the surface in the W, NW and E sectors inside the 
reef edge. Elsewhere deep water extends to the inner reef edge as 
shown by the soundings taken. 

Many small coral heads are evident in the shallower areas. In 
the deep areas, the bottom exhibits highs with small coral heads on 
top, and relatively flat areas both with and without coral heads. Sec- 
tion 3-4 is a good example of these deeper flat areas. Samples of 
coral sand were dredged from clear areas. 

Suwarrow Atoll with depths of over 60 metres is twice the depth 
of Palmerston Atoll. Soundings taken at Suwarrow reveal the bottom 
configuration to be quite different from Palmerston Atoll with a much 
lesser concentration of coral heads and relatively flat clear areas 
in the deeper parts (Figs. 1 and 4). 

Water Characteristics and Tidal Measurements ; 
Temperature and Salinity 

Measurements at 1 station on 21 September 1981 gave water tempera- 
ture of 28.0°C at the surface and down to 24 m, decreasing to 27.9°C 



113 

at 32 m, to 27.8°C at 40 m and to 27.7°C from 40 m to the bottom at 64 
m. Salinity over this depth ranged from 35.51-35.56 /°°. 

Tidal measurements 

Measurements were made using a tide gauge and semi-diurnal tides 
were recorded with a maximum range of 0.69 m and a minimum range of 
0.51 m over a 48 hour period on 20-22 September 1981. Over a 14 day 
period a party on the island using a tide pole recorded readings from 
9 to 92 cm, a range in excess of 0.80 m, but readings of peaks of high 
and lows may have been missed (C. Woodroffe pers. coram.). 

ACKNOWLEDGMENTS 

The author would like to thank Mr. W. deL. Main for assistance in 
the field and Drs. D. E. Hurley and R. A. Pickrill who provided useful 
comments on the manuscript. 

REFERENCES 

Hydrographic Chart Suvorov Islands 1922. 1:36,500, B.A. 1174. 

Irwin, J. & Main, W. deL. 1983. Palmerston Atoll Bathymetry 1:18,000, 
N. Z. Oceanographic Institute Chart, Miscellaneous Series No. 61. 





60 
•H 




Bathymetry 

Depths in metres 

Approximate isobaths r~ " 

Intermediate isobaths ; r- 

Echo sounding traverses 

Topography 

Mean High water mark 

Approximate outer limits /nvywr"'! 

of visible coral "^ 



Fig. 2. Palmerston Atoll Bathymetry (m) of "low" areas. 




Fig. 3. Suwarrow Atoll showing echo sounding traverses. Heavy 
lines show traverses shown in Fig. 4. Track positions 
are approximate only. 




I 



I 







U 
Q) 

13 
C 

O 

tn 

o 
,e 

o 
w 



g 

cd 

I 



so 

•H 







00 
•H 





60 
•H 



a s s s 



SMITHSONIAN INSTITUTION LIBRARIES 



3 9088 01660 9596 



5S m 



^B 



■ 






■ 






r,;> 









-A.V^ 





^^1 




i$M 


H 


^Ikk«mM